Johnson v. Miller

Full title: Davida JOHNSON; Pam Burke; Henry Zittrouer; George Deloach; and George…

Court: United States District Court, S.D. Georgia, Augusta Division

Date published: Sep 12, 1994

864 F. Supp. 1354 (S.D. Ga. 1994)

Opinion

Civ. A. No. 194-008.

September 12, 1994.

A. Rowland Dye, Dye, Tucker, Everitt, Wheale Long, Augusta, GA, Larry H. Chesin, A. Lee Parks, Kirwan, Goger, Chesin Parks, Atlanta, GA, for plaintiffs.

Dennis Robert Dunn, David Frank Walbert, Walbert Herman, Atlanta, GA, for Zell Miller, Pierre Howard and Max Cleland.

Sam George Nicholson, Augusta, GA, Sewell R. Brumby, Office of Legislative Counsel, Atlanta, GA, for Thomas Murphy.

Laughlin McDonald, Neil Bradley, Mary Ellen Wyckoff, ACLU, Atlanta, GA, for Lucious Abrams, Jr., G.L. Avery, Rev., William Gary Chambers, Sr., Rita Valenti and Karen Watson.

Laughlin McDonald, Mary Ellen Wyckoff, ACLU, Atlanta, GA, for Judy Lammers.

Judybeth Greene, Loretta King, Daniel H. Claman, Dept. of Justice — Civ. Rights Div., Donna M. Murphy, Dept. of Justice, Civ. Rights Div., Voting Section, Washington, DC, for U.S.

Doug Teper, pro se.

A. Leon Higginbotham, Jr., New York City, for Congressional Black Caucus amici curiae.

Before EDMONDSON, Circuit Judge, EDENFIELD, Chief District Judge, and BOWEN, District Judge.

MEMORANDUM OPINION AND ORDER

PER CURIAM:

Plaintiffs challenge the constitutionality of Georgia's Eleventh Congressional District and seek an injunction against its further use in congressional elections. Because we find that the district violates the Equal Protection Clause of the Fourteenth Amendment to the United States Constitution, we grant Plaintiffs' request for injunctive relief and will impose a revised plan in keeping with this Memorandum and Order.

INTRODUCTION

In 1993, the Supreme Court recognized a citizen's right under the Equal Protection Clause to challenge a strangely shaped voting district as an impermissible racial gerrymander. See Shaw v. Reno, ___ U.S. ___, 113 S.Ct. 2816, 125 L.Ed.2d 511 (1993). The way was thus cleared for constitutional claims against congressional voting districts in North Carolina, Louisiana, Texas, Florida, and now Georgia.

Southern states have proved fertile ground for Shaw claims, as many of their legislatures labor under the long shadow of the Voting Rights Act of 1965: By that law, certain states or political subdivisions are prohibited from enforcing "any voting qualification or prerequisite to voting, or standard, practice, or procedure with respect to voting" unless they (1) obtain a declaratory judgment from the District Court for the District of Columbia that such change "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," or (2) have submitted the proposed change to the Attorney General and the Attorney General has precleared it. 42 U.S.C. § 1973b-c (1993). This procedure applies to redistricting plans, 28 C.F.R. § 51.13 (1993), and it is intended to police those regions of the United States that had, as of certain dates, maintained voting "tests or devices" serving to disenfranchise minority voters. 42 U.S.C. § 1973b. Many Southern states were so designated. See 28 C.F.R. § 51 (at Appendix) (listing, among other areas, Alabama, Georgia, Louisiana, Mississippi, South Carolina, Texas, Virginia, and sections of North Carolina and Florida).

Consequently, many southern states seek preclearance from the Department of Justice (hereinafter sometimes referred to as "DOJ") before enacting their proposed redistricting plans. Department regulations require DOJ to decide "[w]hether the submitted change has the purpose or will have the effect of denying or abridging the right to vote on account of race, color, or membership in a language minority group"; if the change may be so described, DOJ will not preclear it. 28 C.F.R. § 51.52. In making its determination, DOJ is required to consider the following "important background information":

(1) The extent to which minorities have been denied an equal opportunity to participate meaningfully in the political process in the jurisdiction.

(2) The extent to which minorities have been denied an equal opportunity to influence elections and the decisionmaking of elected officials in the jurisdiction.

(3) The extent to which voting in the jurisdiction is racially polarized and political activities are racially segregated.

(4) The extent to which the voter registration and election participation of minority voters have been adversely affected by present or past discrimination.

Id. § 51.58. It must also consider the following factors specific to the redistricting process:

(a) The extent to which malapportioned districts deny or abridge the right to vote of minority citizens.

(b) The extent to which minority voting strength is reduced by the proposed redistricting.

(c) The extent to which minority concentrations are fragmented among different districts.

(d) The extent to which minorities are over-concentrated in one or more districts.

(e) The extent to which available alternative plans satisfying the jurisdiction's legitimate governmental interests were considered.

(f) The extent to which the plan departs from objective redistricting criteria set by the submitting jurisdiction, ignores other relevant factors such as compactness and contiguity, or displays a configuration that inexplicably disregards available natural or artificial boundaries.

(g) The extent to which the plan is inconsistent with the jurisdiction's stated redistricting standards.

Id. § 51.59.

This litany makes it fairly clear that, by instruction of the United States Congress, racial concerns are DOJ's overriding criterion for approving a redistricting plan.

Shaw v. Reno holds that if a plaintiff shows that racial concerns were the overriding criterion for drafting a redistricting plan, leading to the creation of dramatically irregular district boundaries, that plan is unconstitutional, unless it survives constitutionally strict scrutiny. See Shaw, ___ U.S. at ___ — ___, 113 S.Ct. at 2826-27.

And therein lies the problem.

I. FACTS

Pursuant to the results of the 1980 Decennial Census, the State of Georgia was entitled to ten seats in the United States House of Representatives. Due to population increases revealed by the 1990 Census, that number increased to eleven. This change required the reformatting of Georgia's congressional districts, a task begun by House and Senate reapportionment¹ committees during the 1991 session of the Georgia General Assembly. The task would prove far more onerous than anticipated.

1 Technically, "reapportionment" refers to the allocation of representatives among previously established voting areas, while "redistricting" refers to the reformulation of boundaries for those voting areas. We use the words interchangeably, however, since the legislature, case law, and parties' attorneys do so.

In order to clarify the drafting process, on February 28, 1991 both the House and Senate adopted redistricting guidelines. See Joint Exh. 9, 10. Both versions required public hearings, allowed for submissions by "third parties" outside the Assembly, and provided a list of drafting criteria: single-member districts only, equality of population among districts, contiguous geography, avoiding dilution of minority voting strength, following precinct lines where possible, and compliance with sections 2 and 5 of the Voting Rights Act (hereinafter sometimes referred to as "VRA") 42 U.S.C. § 1973, et seq. Once these requirements were met, drafters could consider maintaining the integrity of political subdivisions, preserving the core of existing districts, and avoiding contests between incumbents.

While the House and Senate surely considered these criteria a realistic tool for drafting reasonable voting districts, and while many of their members were veterans of past redistricting wars, the legislators could not have known what the DOJ would require by way of compliance with sections 2 and 5 of the VRA.

A. The Preclearance Process 1. The Max-Black Plan

One of the "third party" redistricting proposals submitted to the legislature in 1991 would later earn the ominous moniker, "the max-black plan." That plan, created by Ms. Kathleen Wilde, then an attorney with the American Civil Liberties Union ("ACLU") and in her capacity as advocate for the Black Caucus of the Georgia General Assembly, provided for three "majority-minority" congressional districts² in Georgia: the Second, Fifth and Eleventh. Tr. IV, at 70-72, 93; Joint Exh. 4 (version of max-black plan proposed by Reps. Cynthia McKinney and Tyrone Brooks). In contrast, the then existing congressional plan had only one majority-minority district: the Fifth. Joint Exh. 15.

2 That is, a district in which a majority of the population is black.

Ms. Wilde's drive to create three majority-minority districts came, predictably enough, from her clients, including Rep. Tyrone Brooks. Brooks made several illuminating statements during the redistricting debate on the floor of the Georgia House:

This plan [the first redistricting proposal to be submitted to DOJ by the General Assembly] does not come close to the criteria outlined by the Attorney General last summer, when he specifically told the states covered by the Act that wherever possible, you must draw majority black districts, wherever possible.

. . . Obviously, it's possible to draw three; those of us who have been working on the max plan for over a year have really not been concerned about anything other than maximizing our voting strength.

We've not been really concerned about territory; we've not been concerned about voter registration numbers, we've been concerned about population, and black voting age population. When you start playing games with lines and numbers on voter registration and — all of these other things that really don't matter, you are just simply wasting time.

Pltf.Exh. 132, at 40. Ms. Wilde claimed that her purpose "was to draw as many districts as possible consistent with equal opportunity for black citizens in Georgia." Tr. IV, at 71. That is, her plan attempted to secure proportional representation for the black population. Id. at 72. However, despite Ms. Wilde's vague answers to this query at trial, id. at 73-74, it is clear that the max-black plan accounted for virtually all major concentrations of black population in Georgia. Compare DOJ Exh. 1-3 with Joint Exh. 4. Ms. Linda Meggers, Director of Reapportionment Services for the Georgia General Assembly and probably the single most knowledgeable person available on the subject of Georgian redistricting, testified that, without including the heavy black populations present in metropolitan areas, achieving the percentages needed for a majority-minority district was impossible. Tr. I, at 97-98. Ms. Wilde accounted for all such metropolitan populations, including Atlanta, Augusta, Macon and Savannah. Her proffered restraint in not seeking "over-proportional" black representation was disingenuous; a fourth majority-minority district, of equal population to the other ten congressional districts, was probably impossible to create.

The max-black plan further maximized black voting strength by pushing the percentage of black voters within its majority-black districts "as high as possible." Tr. IV, at 72-73 (testimony of Ms. Wilde). Indeed, race was "certainly an overriding factor," id. at 81, 97, in designing Ms. Wilde's district boundaries, as she believed was required by the Voting Rights Act and Thornburg v. Gingles, 478 U.S. 30, 106 S.Ct. 2752, 92 L.Ed.2d 25 (1986). Significantly, she also thought that the proper measure of "compactness," a venerable districting consideration directly relevant to the resulting shape of voting districts, was a "functional" measure, not one grounded in geographical assessments. Tr. IV, at 79. That is, while geography was a "component" of compactness, the more important element was whether people were "able to know what district they're in, know who their representative is and be able to have that person represent them." Id. Ms. Wilde agreed "[t]o some degree" with Rep. Brooks, who told the Georgia House:

So those of us who are promoting the max plan are drawing the lines sometimes in funny ways to cut us in, because we've never really been fully in.

So simply as I say to you today to adopt the max-black congressional plan, I say to you it's not really the appearance of the district; it's the content of the district.

Id. at 95; Pltf.Exh. 132, at 43.

Ms. Wilde testified that she submitted her max-black plan to DOJ several times during the redistricting process. We now turn to the connection between that plan and the one ultimately approved by the Department of Justice.

2. ACLU Advocacy and DOJ Preclearance Requirements

The Court realizes that the motivations of the State of Georgia are the legally relevant motivations in this case, and not those of the ACLU or the DOJ. We consider the behavior of the latter parties, however, because it is inextricably linked to the redistricting decisions of the General Assembly. The actions of the legislature have much less significance when removed from their context.

a. ACLU-DOJ Cooperation

During the redistricting process, Ms. Wilde was in constant contact with both Keith Borders and Thomas Armstrong, the DOJ line attorneys overseeing preclearance of Georgia's redistricting efforts. See Tr. IV, at 39, 231. There were countless communications, including notes, maps, and charts, by phone, mail and facsimile, between Wilde and the DOJ team; those transactions signified close cooperation between Wilde and DOJ during the preclearance process. The Court was presented with a sampling of these communiques, and we find them disturbing.

It is obvious from a review of the materials that Ms. Wilde's relationship with the DOJ Voting Section was informal and familiar; the dynamics were that of peers working together, not of an advocate submitting proposals to higher authorities. See, e.g., Pltf.Exh. 57H, 57I. DOJ was more accessible — and amenable — to the opinions of the ACLU than to those of the Attorney General of the State of Georgia. See Pltf.Exh. 52, 54, 57, 57A-M, 165; Tr. V, at 3-4. It is clear from our proceedings that Ms. Wilde discussed with DOJ lawyers the smallest details of her plan, constantly sending revisions, updates, and data throughout the period from October, 1991 to April, 1992; she occasionally sent documents to DOJ lawyers "per your request." Ms. Wilde worked with DOJ in other ways: During the reapportionment process for Georgia's House districts, DOJ attorney Nancy Sardison told Mark Cohen, the Senior Assistant Attorney General for Georgia, to meet with Ms. Wilde to revise a majority-black House district. Mr. Cohen had presumptuously thought the district satisfactory, but was dutifully informed by Ms. Sardison that Ms. Wilde was "still having some problems with it." Tr. V, at 3.

Contrary to Mr. Armstrong's claims at trial, the max-black proposal was not merely "one of the alternatives [DOJ] considered," and Ms. Wilde was not simply one of various advocates. Her work was of particular importance to DOJ lawyers, whose criteria for and opinions of Georgia's submissions were greatly influenced by Ms. Wilde and her agenda.

Alas, it is true that none of the DOJ attorneys testifying at trial admitted to the influence of Ms. Wilde and her max-black plan on their preclearance deliberations. This Court finds it distressing that Messrs. Borders and Armstrong lacked any significant memory of important elements of the 1991-92 preclearance saga. Both of them — especially Mr. Borders — intimately involved with the redistricting for months, just "don't recall" basic details of either important meetings or the preclearance process. See, e.g., Tr. IV, at 8-51; 145-150. Those in far more peripheral roles had no great difficulty remembering the events central to our inquiry.³ Frankly, based on the factual record and trial testimony, the Court finds Borders' and Armstrong's professed amnesia less than credible. Luckily, the surrounding evidence speaks quite clearly.

3 The Court finds it particularly difficult to believe that Borders simply "doesn't remember" why the Voting Section would have told the General Assembly to extend the Eleventh District all the way down to Chatham County.

Beyond the working relationship noted above, evidence of the link between Ms. Wilde and DOJ's preclearance requirements is provided by simply comparing the max-black plan with DOJ's demands of the General Assembly. While DOJ's policy of minority vote maximization had been in force before the 1991-92 preclearance process,⁴ the effect of Ms. Wilde's efforts on Georgia's redistricting fortunes is evident on the face of the precleared plan. Before comparing that plan to the max-black proposal, we first review the stages of Georgia's quest for DOJ approval.

4 Tr, III, at 24-26 (testimony of Assistant Attorney General for Civil Rights John Dunne). Mr. Armstrong himself testified that:

what we did and what I did specifically was to take a Clorapleth [sic] map of the State of Georgia shaded for race, shaded by minority concentration, and overlay the districts that were drawn by the State of Georgia and see how well those lines adequately reflected black voting strength.

Tr. IV, at 135. This is, essentially, a search for maximization by the crudest means.

b. Round One

After a special session beginning in late August, 1991, the General Assembly submitted a congressional redistricting plan for preclearance on October 1 of that year. The plan, Joint Exhibit 3, contained two majority-minority districts, the Fifth and Eleventh, and a third district, the Second, where blacks comprised an influential but submajority percentage of the voting age public.⁵ See Pltf.Exh. 1, at 8-10.

5 The relevant data for districts 2, 5, and 11 of the first submitted plan are as follows:
----------------------------------------------------------- District Black Pop Black VAP % of Total % of Total ----------------------------------------------------------- 2 39.47 35.37 ----------------------------------------------------------- 5 62.13 57.84 ----------------------------------------------------------- 11 60.63 56.61 -----------------------------------------------------------

This plan, like most redistricting efforts, was the culmination of committee meetings, public hearings, examination of various districting proposals, and many hours spent with an extremely sophisticated computer.⁶ Rep. Robert Hanner, Chairman of the Georgia House reapportionment committee, and Senator Eugene Walker, Chairman of the Georgia Senate reapportionment committee, worked closely with Linda Meggers to hammer out Georgia's first submission. Neither Ms. Meggers, a central figure in construction of all Georgia districting plans submitted to DOJ, nor Senator Wayne Garner, the Majority Leader of the State Senate during the reapportionment process, found any evidence of acts or statements by Rep. Hanner, Senator Walker, or other legislators indicating an intent to discriminate against minority voters. The Court finds no evidence prompting us to doubt that conclusion.⁷

6 Ms. Meggers provided the Court with an enlightening demonstration of state-of-the-art redistricting on her computer, "Herschel." Only Georgians truly understand the depth of respect that is accorded to this equipment by such an appellation. When fed geographic and demographic data (including black voting age population data), it allows the user to map voting districts at the census bloc level with the greatest of ease.

7 The Court has reviewed correspondence from Rep. Cynthia McKinney to John Dunne accusing the Georgia House reapportionment committee of ignoring the Brooks-McKinney max-black House redistricting proposal and only reluctantly allotting computer time to Rep. McKinney's redistricting efforts. Pltf.Exh. 51. Comment letters from the ACLU echo these accusations. Pltf.Exh. 52. Neither the State Defendants nor the Intervenors, however, have presented any further evidence or testimony corroborating Rep. McKinney's claims. Rep. Brooks made no mention of them during his testimony. We also note that he had no difficulty touting the max-black proposal on the floor of the House from the very beginning of the process. See, e.g., Pltf.Exh. 130, 132. The Georgia Senate was equally amenable to Senator Sanford Bishop's advocacy for three majority-black districts. Pltf.Exh. 131. See also Pltf.Exh. 140 (comments of then-Senator Bishop, now a member of the United States House of Representatives, at Georgia House Congressional Committee meeting).
The statements of the Speaker of the House, the Chairpersons of the House and Senate committees, and Linda Meggers all reveal a legislature driven to satisfy the demands of DOJ, and contain no indication of efforts to suppress the black vote. Thomas Armstrong could not recall seeing any evidence of discriminatory motives on the part of the Georgia legislature. Tr. IV, at 145-50.

The first submitted plan did not garner universal support within the General Assembly; a version of the max-black plan had been proposed in the Georgia House, but could not attract enough votes for stretching the new Eleventh District to Savannah. Tr. III, at 248-50; Pltf.Exh. 6A-B. More importantly, the plan garnered no support from DOJ, which rejected it by letter dated January 21, 1992. Pltf.Exh. 2.

In that letter, John Dunne, then Assistant Attorney General for Civil Rights, noted a "concern" that the Georgia legislative leadership had decided to "limit" black voting potential to two majority-minority districts. Throughout the preclearance process, from this first objection letter to the final submission, DOJ relied on versions of the max-black plan to argue that three majority-minority districts could indeed be squeezed out of the Georgia countryside. Tr. IV, at 115 (testimony of T. Armstrong). Ms. Wilde's triumph of demographic manipulation became DOJ's guiding light. Though DOJ denied that the max-black plan was the "benchmark" against which Georgia's efforts were compared, its role as such soon became obvious to the General Assembly, and is now obvious to this Court.

Mr. Dunne also observed that, while the submitted plan properly utilized black voting potential in Atlanta and east-central Georgia, it did not "recognize" concentrations of minorities in the south-west region of the state. He then assured Mark Cohen that "[s]ection 5 considerations certainly do not dictate that the state adopt any particular configuration," but went on to "note" the "exclusion" from the Eleventh District of black populations in adjacent Baldwin County. Pltf.Exh. 2.

Mr. Dunne ultimately found himself unable to conclude that the Eleventh District, 60.63% black, with a black voting age population of 56.61%, and using nearly unpopulated land bridges to rope in black concentrations in DeKalb and Richmond counties, satisfied the requirements of the Voting Rights Act. Apparently finding that this district — and the plan generally — "ha[d] the purpose or will have the effect of denying or abridging the right to vote on account of race," Mr. Dunne refused to preclear the submitted plan. Id.

c. Round Two

No one in the General Assembly doubted that any revised submission must include the changes "suggested" by DOJ. Though counsel for the United States objected to Plaintiffs' "characterization that the Justice Department 'suggested' things," Tr. IV, at 120, it is disingenuous to submit that DOJ's objections were anything less than implicit commands. No one in the General Assembly — including the Black Caucus — thought so, and DOJ lawyers did nothing to dissuade legislators of that notion.⁸

A few prescient souls were already convinced that any successful Georgia submission would have to "have the highest percentage of black population that we could get, irregardless [sic] of where we have to go." Tr. III, at 210. Senator Garner was convinced that, though "distasteful," "to get a plan passed the Justice Department and get us out of here and on to these elections, that district is going to have to go to Savannah." Id. Having convinced the State Senate of this political reality, that body passed a plan extending the Eleventh District to Chatham County. Id. at 211-12; DOJ Exh. 76. The House, however, refusing to tack Savannah to the Eleventh until DOJ forced them to do it, did not pass the Senate bill. Tr. III, at 213, 250. The final version, submitted on March 3, 1992, did not contain the Savannah extension. DOJ Exh. 76.

It did, however, contain a Second District in which black voting strength had been increased to 45.01%, and a Fifth and Eleventh District where blacks comprised close to 60% of the voting age population.⁹ Pltf.Exh. 3A. In the Eleventh, faithful to DOJ's first objection letter, the General Assembly split Baldwin County to "recognize" the black populations that had previously been "excluded" from that district. Many black representatives voted for the plan, more in the Senate than in the House. Pltf.Exh. 6A-B.

9 The relevant data for districts 2, 5, and 11 of the second submitted plan follows:
----------------------------------------------------- District Black Pop Black VAP % of Total % of Total ----------------------------------------------------- 2 49.15 45.01 ----------------------------------------------------- 5 62.27 57.47 ----------------------------------------------------- 11 61.86 57.97 -----------------------------------------------------

Predictably, DOJ rejected the plan, since, after all, someone — here the Senate relying on versions of the max-black proposal — had come up with a scheme for packing even more minority voting strength into the Second, Fifth and Eleventh Districts. And so, despite the fact that the second submission addressed the criticisms levelled at the first, DOJ sent the plan back to the General Assembly yet again, and this time with new demands.

Since it was geographically possible to extend the Eleventh District to Savannah, Mr. Dunne, by letter dated March 20, 1992, made it a requirement for preclearance:

[T]he submitted plan minimizes the electoral potential of large concentrations of black population in several areas of the state. Specifically, we note that alternatives, including one adopted by the Senate, included a large number of black voters from Screven, Effingham and Chatham Counties in the 11th Congressional district. However, due to unyielding efforts on behalf of House members, this configuration was abandoned and no legitimate reason has been suggested to explain the exclusion of the second largest concentration of blacks in the state from a majority black Congressional District.

Pltf.Exh. 4, at 3. The first rejection letter never mentioned this extension to Screven, Effingham and Chatham Counties.¹⁰

10 DOJ attorneys had suggested an extension to Chatham County during a meeting in Washington after receipt of the first objection letter and submission of Georgia's second plan. Tr. V, at 10-14 (testimony of Mark Cohen). See also Tr. IV, at 192 (testimony of Lt. Gov. Pierre Howard).

Mr. Dunne then noted that "[a]lthough the submitted plan has increased the black percentage in the 2d Congressional District, it continues the exclusion of large black population concentrations in areas such as Meriwether, Houston, and Bibb Counties from this district." Id. Of course, DOJ was aware that the only way to transfer the dense black population of Bibb County from the Eleventh District to the Second District and still maintain the Eleventh as a majority-minority district was to extend the Eleventh to Savannah, where dense black populations in that city would make up for the loss of black voters in Bibb County. Georgia Attorney General Michael Bowers made the observation himself in the second plan submitted to DOJ. Attorney General Bowers also explained why the state had declined to follow that course:

The reason the General Assembly chose not to adopt a configuration for the 2nd District which would have altered the 11th District in this fashion is two-fold. First, if the 11th District extends from the Atlanta area in DeKalb County into the metropolitan areas of Richmond, Bibb, and Chatham Counties, a candidate to be successful will have to run in four major media markets in Georgia. This would likely present a significant problem for any minority candidate in the 11th District, as recognized even by the proponents of the "Brooks-McKinney" [max-black] plan. . . . Second, it is clear that even if the population percentage is increased several more points in the 2nd District . . . the 2nd District will still be no more than an influence district for minority voters. Black voter registration in this area has been historically low. . . . In addition, the extension of the 2nd District into Bibb County and the corresponding extension of the 11th District into Chatham County, with all of the necessary attendant changes, violate all reasonable standards of compactness and contiguity.

Pltf.Exh. 3A, at 8.

The Macon/Savannah "trade" was the linchpin of Ms. Wilde's max-black plan and all the versions of it submitted to the Assembly; Ms. Wilde had been marketing it since at least October 15, 1991, describing the switch to DOJ as the "key to drawing a third black district." Pltf.Exh. 57 (facsimile from Ms. Wilde to Sheila Delaney, DOJ Voting Section). It was the centerpiece of her advocacy for the General Assembly Black Caucus:

MR. PARKS: The advocacy that you met [sic] with the Department of Justice was to move Bibb County — the black population of Bibb County to the Second?

MS. WILDE: That's right.

MR. PARKS: And replace them by dropping the Eleventh all the way down to Savannah?

MS. WILDE: That's correct.

MR. PARKS: And excise out of Savannah black population?

MS. WILDE: I took majority black precincts in Savannah and put them in my district, which is how you draw a majority black district.

MR. PARKS: And intentionally excluded white population?

MS. WILDE: Again, I did that because that is how you draw a majority black district.

Tr. IV, at 83. See also id. at 82.

It is from Ms. Wilde's plan and its progeny that DOJ lifted the concept for its own objection letters. It "was the evidence relied upon for the three minority district proposition." Tr. IV, at 115 (testimony of T. Armstrong). Armstrong refused to admit that Mr. Dunne's second objection letter was beholden to the max-black plan, observing that Ms. Wilde's plan was but one of many alternatives presented to DOJ. Id. at 116. The Court notes, however, that the necessary kernel of a viable three-district plan — the Macon/Savannah trade — originated with Ms. Wilde. Her plan was premised upon it, and was the foundation of the first three-district plan proposed in the General Assembly. Other plans presenting three majority-black districts were in turn based on that one. Any such proposal reviewed by DOJ switched Macon from the Eleventh to the Second District and drew Savannah into the Eleventh, and it owed that crucial element to Ms. Wilde's original version.

d. Round Three

It was now clear to the General Assembly that preclearance would not be forthcoming without adopting this raison d'être of the max-black proposals.¹¹ This goal dominated the creation of the third Georgia submission. Linda Meggers admitted that, when drafting the proposal, the max-black plan was her "benchmark," Tr. I, at 65, and that she "fine tuned" the plan to "get to those numbers." Id. at 102.

11 We note that the State of Georgia did not choose to seek a declaratory judgment from the D.C. District Court when it became obvious that DOJ would accept nothing less than abject surrender to its maximization agenda. Mark Cohen testified that since the circumstances of the 1991-92 redistricting were similar to those of 1981-82, such an effort would have been unsuccessful. In the 1981-82 process, Georgia sought a declaratory judgment that the state had gone far enough in strengthening black voting power, and was not required to ratchet up the black voting age population of the 5th congressional district to 65%. As here, the Georgia Senate had passed a plan acquiescing to DOJ's demands, and the House had refused to endorse it. The D.C. court held that Georgia had failed to satisfy the requirements of section 5. See Busbee v. Smith, 549 F. Supp. 494 (D.D.C. 1982), aff'd, 459 U.S. 1166, 103 S.Ct. 809, 74 L.Ed.2d 1010 (1983).
Mr. Cohen's conclusion was buttressed by Sandra Coleman, Deputy Chief of the DOJ Voting Section, who noted these similarities during the second meeting between Georgia and DOJ officials in Washington, D.C. Add to this the cost of litigating in D.C. and the time pressures on the Assembly, and it becomes clear that a declaratory judgment action was not likely.

The result of these efforts was a congressional districting plan with three majority-minority districts: the Second, with 52.33% black voting age population, the Fifth, with 57.47%, and the Eleventh, with 60.36%.¹² The plan passed by one vote — Speaker Thomas Murphy "held [his] nose and shut [his] eyes and put one elbow in one ear and voted for it and passed it to get something to keep the courts from doing it." Dep. of Spkr. Murphy, at 31. Ms. Meggers came as close as she could to copying the max-black percentages. She was prevented from exact duplication by the need to construct eight other districts around the majority-minority districts, and her efforts to "clean up" the max-black plan to minimize the number of split counties. Compare Joint Exh. 1 with Joint Exh. 4. Ms. Wilde herself thought that Georgia's third plan was a fairly successful recreation of her proposal:

12 The relevant data for districts 2, 5, and 11 of the third proposed plan follows:
-------------------------------------------------------- District Black Pop Black VAP % of Total % of Total -------------------------------------------------------- 2 56.63 52.33 -------------------------------------------------------- 5 62.27 57.47 -------------------------------------------------------- 11 64.07 60.36 --------------------------------------------------------

MR. PARKS: [T]he result was the creation of three minority districts that achieved the percentage benchmarks that the Max Black plan advocated?

MS. WILDE: Yes, that's what the State of Georgia decided to draw.

Tr. IV, at 100. Rep. Brooks agreed. See Tr. IV, at 229.

The latest redistricting plan bore all the signs of DOJ's involvement: The black population of Meriwether County was gouged out of the Third District and attached to the Second District by the narrowest of land bridges; Effingham and Chatham Counties were split to make way for the Savannah extension, which itself split the City of Savannah; and the plan as a whole split 26 counties, 23 more than the existing congressional districts. Joint Exh. 1. The dense population centers of the approved Eleventh District were all majority-black, all at the periphery of the district, and in the case of Atlanta, Augusta and Savannah, all tied to a sparsely populated rural core by even less populated land bridges.¹³ Extending from Atlanta to the Atlantic, the Eleventh covered 6,784.2 square miles, splitting eight counties and five municipalities along the way.

13 We further explore the characteristics of the Eleventh District in our discussion of strict scrutiny, infra.

DOJ precleared the plan, which was hand delivered to DOJ by Mark Cohen. During his visit, Keith Borders opined that even this plan did not go far enough, voicing objections that Ms. Wilde and Rep. Brooks were even then faxing to the Voting Section. On April 2, 1992, however, Mr. Cohen was finally able to leave Washington with a letter of preclearance from Mr. Dunne. Pltf.Exh. 6.

B. The Ramifications of DOJ Involvement

1. DOJ Methods a. The Use of Informants

During our hearings it became clear that the Department of Justice had cultivated a number of partisan "informants" within the ranks of the Georgia legislature, including at least one State Senator — a congressional candidate no less — and an aide to Lieutenant Governor Howard. See Response of United States To Plaintiffs' Statement Concerning Confidential Informants, July 7, 1994. DOJ regularly received from them information on the General Assembly's redistricting sessions. DOJ apparently read federal regulations as condoning this behavior, but misconstrued the spirit of those provisions. See 28 C.F.R. § 51.29. They were intended to facilitate outside notification to DOJ of impending changes in voting procedure in jurisdictions covered by section 5, since DOJ could not closely monitor all such areas on a continual basis. The regulations require DOJ to respect requests for anonymity, in order to encourage individuals to report. The events in this case do not reflect the regulations' intended purpose. Politicians and other parties in interest found a discrete forum. Here, "whistleblowers" became "secret agents." They reported to DOJ throughout the section 5 review process. DOJ used that information even to question the integrity of State legislators who could not know their accusers.

Keith Borders testified that "allegations are brought to our attention and then we go back and ask individuals about those allegations." Tr. IV, at 9. One of DOJ's informants informed Mr. Borders that State Senator Eugene Walker was a "quintessential Uncle Tom," and "the worst friend of blacks in Georgia." Tr. IV, at 27. During the first meeting between a Georgia delegation and Voting Section lawyers, there was a confrontation involving Borders and Walker, although Borders "doesn't recall" whether he challenged Senator Walker with those unattributable comments. However, either these anonymous "allegations" or some other suggestion so offended Senator Walker that he refused to attend further meetings with DOJ staff.

DOJ's agents provided an irrebuttable source of intelligence of which the General Assembly was completely unaware. We find this practice disturbing.

b. The Use of Alternative Plans

Thomas Armstrong testified that DOJ used alternative plans submitted by special interests as means with which to question the configuration of Georgia's submitted plans, i.e., if an alternative proposal had found some clever means of further boosting African-American voting strength in Georgia, why hadn't the Assembly adopted it? See, e.g., Tr. IV, at 140-41. For DOJ, if these alternative plans had discovered heretofore untapped wells of racial voting power absent from the submitted plan, the inescapable conclusion was that the State's proffered reasons for the submitted configurations were "pretextual" ones.

There is a problem with this position: It apparently did not occur to DOJ that increased "recognition" of minority voting strength, while perhaps admirable, is properly tempered with other districting considerations. The ACLU and Reps. Brooks and McKinney were concerned only with racial percentages. They could afford such a narrow focus. The General Assembly leadership was concerned with passing redistricting legislation affecting all Georgians, and contended with numerous factors racial, political, economic, and personal. As the lawmaking body for the entire state, the General Assembly could not afford such a narrow focus. Consequently, its submissions reflected many influences, and the ACLU's reflected only race. Of course DOJ could then compare Georgia's plans to the max-black plan and find the latter a more effective means of magnifying the minority vote. The max-black plan reflected nothing but its drafters' concerns: race and technical contiguity. DOJ's mistake was in assuming that Georgia's plan must be so conceived. As one legislator quipped:

[I]t's very difficult, it seems, to get them to realize that maybe there's a little bit more to reapportionment than black and white or minority areas; that there should be some alignment between the areas so they can work together for the next ten years.

Pltf.Exh. 143. Difficult indeed.

2. ACLU Involvement

It is unclear whether DOJ's maximization policy was driven more by Ms. Wilde's advocacy or DOJ's own misguided reading of the Voting Rights Act. This much, however, is clear: the close working relationship between Ms. Wilde and the Voting Section, the repetition of Ms. Wilde's ideas in Mr. Dunne's objection letters, and the slow convergence of size and shape between the max-black plan and the plan DOJ finally precleared, bespeak a direct link between the max-black plan formulated by the ACLU and the preclearance requirements imposed by DOJ.

Succinctly put, the considerable influence of ACLU advocacy on the voting rights decisions of the United States Attorney General is an embarrassment. A poignant example: in a notable faux pas, DOJ's second objection letter arrived at the Office of the Attorney General of Georgia only after members of the Georgia Black Caucus were already discussing it with the press. DOJ had notified the ACLU of its impending objection; the ACLU then notified the Black Caucus. This unfortunate spate of gossip created the impression that the ACLU and the Black Caucus wielded significant influence with DOJ's Civil Rights Division and significant control over Georgia's redistricting efforts. The State's leaders were understandably nonplussed. The ACLU was exuberant. Georgia officials and citizens were mystified.

We are simply troubled by the result. It is surprising that the Department of Justice was so blind to this impropriety, especially in a role as sensitive as that of preserving the fundamental right to vote. That right is vital to citizens of a democracy, and of poignant importance to African-Americans. For many blacks Jim Crow is living memory,¹⁴ and the presence of black luminaries and ordinary citizens at our hearings is a testament to their concern. This Court does not underestimate the emotional investment in our decision of blacks still resisting the vestiges of racial discrimination,¹⁵ and we have been placed in the unenviable position of depriving black citizens of a privilege the Justice Department never had the right to grant: maximization of the black vote, whatever the cost.

14 See South Carolina v. Katzenbach, 383 U.S. 301, 310-15, 86 S.Ct. 803, 809-12, 15 L.Ed.2d 769 (1965) for a review of modern state government efforts to deny blacks the ability to vote.

15 We acknowledge in this regard the amicus contribution of Judge A. Leon Higginbotham, Jr., Special Counsel to the Congressional Black Caucus. Various of Judge Higginbotham's comments echoed those of ex-Labor Commissioner Albert Scott, who testified to the remains of racial bigotry in rural Georgia. Compare Amicus Brief at 16 (asserting that Georgia Supreme Court Justice Robert Bentam won elections in 1984 and 1990 "with a strategy of obscuring the fact that he is black") with Tr. VI, at 79 (testifying that some white political candidates use the tactic of specifically advertising to rural voters that their opponents are black).

* * * * * *

Elections were held under the new congressional redistricting plan on November 4, 1992. Black candidates were elected to the United States Congress from all three majority-minority districts. Cynthia McKinney, running in the Eleventh, became Georgia's first black congresswoman. On January 13, 1994, Plaintiffs filed this action.

II. LAW A. Shaw v. Reno

Though initially daunting, the Supreme Court's opinion in Shaw v. Reno, ___ U.S. ___, 113 S.Ct. 2816, 125 L.Ed.2d 511 (1993), is a bit less than a riddle wrapped in an enigma. We begin with some general observations, and then continue with an analysis of the Eleventh District.

The majority opinion in Shaw is premised on the uncontroversial proposition that race-based redistricting should be treated like any other race-based legislation. From there, the argument is straightforward: (1) racial classifications, per the Fourteenth Amendment, are "presumptively invalid, and can be upheld only upon an extraordinary justification." Brown v. Board of Education, 347 U.S. 483, 74 S.Ct. 686, 98 L.Ed. 873 (1954); (2) in light of the Voting Rights Act, race is obviously a valid consideration in redistricting, but a voting district that is so beholden to racial concerns that it is inexplicable on other grounds becomes, ipso facto, a "racial classification;" (3) if a district may be so described, it must have an "extraordinary justification," that is, it must survive constitutionally strict scrutiny, in order to be upheld.

The above analysis begs the question: How to determine if a district is so founded upon race that it warrants constitutional notice? By examining the shape of the district itself: if its contours are so contorted as to permit no other conclusion than that it was drawn along racial lines, the Supreme Court says those lines are clear — if circumstantial — evidence that the legislature "purposefully distinguish[ed] between voters on the basis of race." Shaw, ___ U.S. at ___, 113 S.Ct. at 2826. This analysis is made a bit easier by the existence of defined and recognized "traditional districting principles" that influence nearly all redistricting efforts. The nonexclusive list includes contiguity, compactness, protecting the integrity of political subdivisions, cognizance of "communities of interest," negotiating geographic barriers, and protecting incumbents. If the shape of the district advertises "disregard" for these considerations in favor of race-based line drawing, the district is a racial classification and subject to strict scrutiny.¹⁶

16 The Supreme Court takes a reverse tack in describing the significance of traditional districting principles to a Shaw claim:

We emphasize that these criteria are important not because they are constitutionally required — they are not, cf. Gaffney v. Cummings, [ 412 U.S. 735, 752 n. 18, 93 S.Ct. 2321, 2331 n. 18, 37 L.Ed.2d 298 (1973)] — but because they are objective factors that may serve to defeat a claim that a district has been gerrymandered on racial lines.

Id. at ___, 113 S.Ct. at 2827 (citation omitted).

There is an important, if subtle, distinction to make here. Under Shaw, the shape of the district is "objective" evidence of the legislature's intent in drafting that district. Id. Discovering discriminatory intent is necessary to a successful Equal Protection claim. Washington v. Davis, 426 U.S. 229, 96 S.Ct. 2040, 48 L.Ed.2d 597 (1976); Mobile v. Bolden, 446 U.S. 55, 100 S.Ct. 1490, 64 L.Ed.2d 47 (1980) (discussing "discriminatory purpose" in Fifteenth Amendment cases). The shape of the district is not a "threshold" inquiry preceding an exploration of the motives of the legislature. That is, the court does not assess, on first principles, whether the district looks "bizarre," and then, if it does, proceed to an Equal Protection analysis. A determination that the district looks dramatically irregular is the beginning of an Equal Protection analysis; the strange district is the "smoking gun," revealing the racial intent needed for an Equal Protection claim.

We take pains to clarify this point because some have misinterpreted Justice O'Connor's observation that "reapportionment is one area in which appearances do matter." Shaw, ___ U.S. at ___, 113 S.Ct. at 2827. Yes, appearances do matter, but in two extremely particular ways: (1) as a proxy for direct evidence of the legislature's intentions, and (2) as a harm adequate to meet the constitutional criteria for standing to sue. The former idea is not a wholly new one. As the High Court notes, there are intellectual precursors to the Shaw opinion making it a logical extension of past practice. See Shaw, ___ U.S. at ___ — ___, 113 S.Ct. at 2825-26 (reviewing cases). See also Rogers v. Lodge, 458 U.S. 613, 102 S.Ct. 3272, 73 L.Ed.2d 1012 (1982) (deferring to district court finding that sum of evidence supports inference that voting scheme was maintained for discriminatory purposes). The latter concept, however, is a bit of an innovation.

1. Standing to Bring a Shaw Claim

In both Shaw and the instant case, the plaintiffs suffered no individual harm; the 1992 congressional redistricting plans had no adverse consequences for these white voters. Under the Supreme Court's most recent pronouncements, this lack of concrete, individual harm would deny them standing to sue. See, e.g., Lujan v. Defenders of Wildlife, ___ U.S. ___, ___, 112 S.Ct. 2130, 2136, 119 L.Ed.2d 351 (1992). Shaw, however, implicitly recognizes a different kind of harm, distinct from vote dilution and sufficient to state a "cognizable" Equal Protection claim: "Classifying citizens by race, as we have said, threatens special harms that are not present in our vote dilution cases. It therefore warrants different analysis." Shaw, ___ U.S. at ___, 113 S.Ct. at 2828. Those harms are systemic ones, rooted in social perception of statesanctioned racial classifications:

Justice Souter [in dissent] apparently believes that racial gerrymandering is harmless unless it dilutes a racial group's voting strength. . . . As we have explained, however, reapportionment legislation that cannot be understood as anything other than an effort to classify and separate voters by race injures voters in other ways. It reinforces racial stereotypes and threatens to undermine our system of representative democracy by signalling to elected officials that they represent a particular racial group rather than their constituency as a whole. . . . Justice Souter does not adequately explain why these harms are not cognizable under the Fourteenth Amendment.

Id. It is this distinct kind of damage that enables the Court to distinguish Shaw from a group of white plaintiffs' failed vote dilution claim in United Jewish Organizations, et al. v. Carey, 430 U.S. 144, 97 S.Ct. 996, 51 L.Ed.2d 229 (1977) (hereinafter "UJO"); in that case, the districts were regularly shaped. See id. at 167-68, 97 S.Ct. at 1010-11 (opinion of White, J.). Shaw explicitly reserves any thoughts on whether the deliberate creation of a majority-minority district, "without more," — meaning without a dramatically irregular shape — gives rise to an Equal Protection claim. Shaw, ___ U.S. at ___, 113 S.Ct. at 2828.¹⁷ Shaw v. Reno's expanded notion of harm liberalizes the standing requirement. The Eastern District of North Carolina, analogizing principally from Regents of the Univ. of Calif. v. Bakke, 438 U.S. 265, 281, 98 S.Ct. 2733, 2743, 57 L.Ed.2d 750 (1978), articulates the standing requirement for a Shaw claim as a showing that a redistricting plan has assigned a potential plaintiff to a district at least in part because of her race. Shaw v. Hunt, 861 F. Supp. 408, 426 (E.D.N.C. 1994). The North Carolina court pursues an exhaustive discussion of Shaw v. Reno's implications for the standing requirement, and we do not repeat it here. We feel that even that court's interpretation may not be as expansive as Shaw v. Reno intended, but are content to note that standing in this case is not in doubt: Plaintiffs occupy a position virtually identical to those challenging North Carolina's congressional redistricting plan in Shaw v. Reno, and in that case standing was not in doubt.

17 The court in Shaw v. Hunt, 861 F. Supp. 408, 432 (E.D.N.C. 1994), dealing with the Supreme Court's remand, wonders about this reluctance, positing that perhaps the Supreme Court sought to avoid overturning its decision in UJO by stopping short of answering that question.
The distinction is more principled, however. The special harm that is the crux of Shaw v. Reno, the offense to "principles of racial equality" generated by blatantly manipulated voting districts, simply was not present in UJO, a case involving vote dilution, not district appearance. The issue of Equal Protection claims against normally shaped districts was beyond the scope of the Shaw opinion. For our part, we note that a compact majority-black voting district could not be challenged as a racial "gerrymander"; if the district is sufficiently compact, then presumably its drafters were able to accumulate enough black voters without carefully picking their way around white populations in search of black ones. Without that kind of manipulation, there is no "gerrymandering" to speak of — the state simply made a discrete voting district out of a discrete community. The Shaw v. Reno majority makes a similar point. See id. at ___, ___, 113 S.Ct. at 2826, 2828. See also Wright v. Rockefeller, 376 U.S. 52, 53-58, 84 S.Ct. 603, 603-06, 11 L.Ed.2d 512 (1964).

2. The Required Level of Racial Motivation in Redistricting

There has been some debate over the necessary prominence of race in legislative deliberations before it can be found that the redistricting plan is "unexplainable on grounds other than race," and thus subject to strict scrutiny. There are three possible solutions. Defendants in the instant case, along with some district courts, argue that race must have been the sole motivation behind a particular district shape before strict scrutiny is appropriate. Defendants cite language from Shaw in support: "[strict scrutiny is applied if the plan] is so irrational on its face that it can be understood only as an effort to segregate voters . . . because of their race[.]" Shaw, ___ U.S. at ___, 113 S.Ct. at 2832. See also Bridgeport Coalition v. City of Bridgeport, 26 F.3d 271 (2d Cir. 1994) (finding that a lower court order "did not transgress Shaw because it did not instruct the City Council to redistrict solely on racial grounds) (emphasis added); DeWitt v. Wilson, 856 F. Supp. 1409 (E.D.Cal. 1994) (stating, for example, "Shaw held when districts are drawn in such an extremely irregular fashion as to be unexplainable, other than being based solely on race, a claim under the Equal Protection Clause for racial gerrymandering can be stated.") (emphasis added).

A second school teaches that race need only have been a recognizable factor — not the sole or dominant one — before a redistricting plan is constitutionally suspect. See Hays v. Louisiana, 839 F. Supp. 1188, 1202, 1214 (W.D.La. 1993) (majority opinion and concurring opinion) (hereinafter "Hays I"). Shaw v. Hunt similarly holds that, in keeping with prior Equal Protection precedents, districts should be subject to strict scrutiny merely upon a showing that the state's use of race was deliberate.¹⁸ Shaw v. Hunt, supra, 861 F. Supp. at 429, 431. This test

18 It is not completely clear exactly what test Shaw v. Hunt espouses. The language of the opinion seems occasionally contradictory. See Majority Opinion, supra, 861 F. Supp. at 431.

is necessarily met by proof that the plan's lines were deliberately drawn so as to create one or more districts in which a particular racial group is a majority, even if factors other than race are shown to have played a significant role in the precise location and shape of those districts.

Id., 861 F. Supp. at 431. Specific intent is needed; the test is not met by demonstrating merely that the legislature was aware of a districting plan's racially discriminatory impact. In order to avoid condemning to "constitutional invalidity" all majority-minority voting districts drawn pursuant to Voting Rights Act requirements, the court states that while this test will subject practically all majority-black districts to strict scrutiny, the strict scrutiny standard should be applied "in a way that is sensitive not only to the state legislatures' statutory obligation [to give effect to minority voting potential under the VRA], but also to the special compromises that they must make in order to pass plans that draw such districts." Id., 861 F. Supp. at 434.

There is an analytical problem with applying strict scrutiny while remaining "sensitive" to the plight of legislatures. Essentially, strict scrutiny tempered to accommodate legislative realities is not "strict scrutiny;" it is some lower standard designed to accommodate legislative realities. We feel that a better approach is to remove completely from constitutional suspicion those VRA-mandated redistricting plans that, as Plaintiffs say, "do it right." The proper interpretation of Shaw v. Reno's "race-based" requirement is that, in order to invoke strict scrutiny, it must be shown that race was the substantial or motivating ¹⁹ consideration in creation of the district in question. That term requires that the legislature (a) was consciously influenced by race, and (b) while other redistricting considerations may also have consciously influenced the district shape, race was the overriding, predominant force determining the lines of the district. If race, however deliberately used, was one factor among many of equal or greater significance to the drafters, the plan is not a racial gerrymander/racial classification subject to strict scrutiny.

19 We intend "motivating" in the sense that race was the most prominent element driving the legislature's planning, not in the sense of one motivation among others of equal strength propelling the process.

We arrive on this middle ground after considering fundamental aspects of Equal Protection jurisprudence. After Washington v. Davis, intent is an indispensable element of successful discrimination claims under the Fourteenth Amendment. The plaintiff must show some discriminatory purpose impelling the legislature before strict scrutiny will be applied to the law in question. See 426 U.S. at 240, 96 S.Ct. at 2047-48. The Supreme Court has expounded on this intent requirement, holding that the Fourteenth Amendment is violated when race is a "substantial" or "motivating" factor in legislative decision-making. See Arlington Heights v. Metropolitan Housing, 429 U.S. 252, 265-66, 97 S.Ct. 555, 563, 50 L.Ed.2d 450 (1977) (clarifying Washington v. Davis); City Schl. Dist. Bd. of Educ. v. Doyle, 429 U.S. 274, 287, 97 S.Ct. 568, 576, 50 L.Ed.2d 471 (1977) (adding word "substantial"). Decisions such as Arlington Heights are vital to properly reading Shaw v. Reno:

Davis does not require a plaintiff to prove that the challenged action rested solely on racially discriminatory purposes. Rarely can it be said that a legislature or administrative body operating under a broad mandate made a decision motivated solely by a single concern, or even that a particular purpose was the "dominant" or "primary" one. . . . When there is a proof that a discriminatory purpose has been a motivating factor in the decision, . . . judicial deference is no longer justified.

Arlington Heights, 429 U.S. at 265-66, 97 S.Ct. at 563.

In the context of Equal Protection claims based on voting districts, the VRA is the kind of "broad mandate" intended by the Court under the Arlington Heights standard. Further, the phrase "unexplainable on grounds other than race," which the Shaw v. Reno majority frequently mentioned, is from the Arlington Heights opinion: "Sometimes a clear pattern, unexplainable on grounds other than race, emerges from the effect of the state action even when the governing legislation appears neutral on its face." Arlington Heights, 429 U.S. at 266, 97 S.Ct. at 564. See Shaw v. Reno, ___ U.S. at ___, 113 S.Ct. at 2825.

The intent requirement extends to voting cases, e.g., Rogers v. Lodge, 458 U.S. 613, 102 S.Ct. 3272, 73 L.Ed.2d 1012 (1982), and Shaw discusses certain voting precedents at length. Shaw v. Reno, ___ U.S. at ___ — ___, 113 S.Ct. at 2825-27 (reviewing Guinn v. United States, 238 U.S. 347, 35 S.Ct. 926, 59 L.Ed. 1340 (1915), Gomillion v. Lightfoot, 364 U.S. 339, 81 S.Ct. 125, 5 L.Ed.2d 110 (1960), and Wright v. Rockefeller, 376 U.S. 52, 84 S.Ct. 603, 11 L.Ed.2d 512 (1964)). These cases clarify Shaw. For example, in Wright, minority plaintiffs brought suit under the Fourteenth and Fifteenth Amendments, claiming that several congressional districts in New York City were "irrationally" shaped and contrived to segregate voters on the basis of race. Wright, 376 U.S. at 53, 84 S.Ct. at 603-04. A three-judge panel, presented with maps, statistics and other indirect evidence, dismissed the complaint, finding that the plaintiffs had "failed to prove that the New York Legislature was either motivated by racial considerations or in fact drew the districts on racial lines." Id. at 56, 84 S.Ct. at 605. The Supreme Court affirmed the dismissal, observing that while "[i]t may be true . . . that there was evidence which could have supported inferences that racial considerations might have moved the state legislature, . . . we agree [with the lower court] that there also was evidence to support [a] finding that the contrary inference was 'equally, or more, persuasive.'" Id. at 56-57, 84 S.Ct. at 605. That is, the circumstantial evidence did not show that the New York congressional lines were influenced predominantly by race — other, equally plausible "inferences" could be drawn from the maps and charts.

So it is in Shaw v. Reno, but the standard is a bit tougher. In Wright, the congressional districts were upheld, but without the patina of legitimacy cast by the VRA, and despite "[defendant] New York's frank concession that it is not possible to say 'that race is irrelevant to redistricting.'" Id. 376 U.S. at 61, 84 S.Ct. at 608 (Douglas, J., dissenting). Shaw concedes that latter point, ___ U.S. at ___, 113 S.Ct. at 2826, thus elevating the finding needed for a successful Equal Protection challenge. When we read Arlington Heights and progeny in conjunction with Wright, we surmise the following: Where it cannot be shown that race was the "substantial" or "motivating" factor behind a voting district by demonstrating that racial concerns are the only ones plausibly to be inferred²⁰ from the district's lines, there is no valid Equal Protection claim.

20 We consider the role of direct evidence in section II.B, infra.

In contrast to Hays I and Shaw v. Hunt, race can be a factor for the legislature, meaning one factor given no more prominence than various others, without triggering strict scrutiny. The legislature may intentionally consider race in redistricting — and even alter the occasional line in keeping with that consideration — without incurring constitutional review. It is the abuse of that privilege, exposed to the world via perverse district shapes, "unexplainable on grounds other than race," that sparks further examination.²¹ We think the race-saturated requirements of the Voting Rights Act require this result. Without such an allowance, and considering the present attitude of the Department of Justice, every state and local government in the United States subject to section 5 of the Voting Rights Act will be buried under Shaw litigation every time it passes a redistricting plan.²² Since each plan must survive strict scrutiny — a factual inquiry — many lawsuits will result in full trials. We are reluctant to conclude that the Supreme Court intended voting rights litigation to sweep the country at ten-year intervals.

21 The analytical distinction here is similar to that made by Justice Powell in the affirmative action context. See Regents v. Bakke, 438 U.S. 265, 317-18, 98 S.Ct. 2733, 2762-63, 57 L.Ed.2d 750 (1977).

22 Cases such as the instant one show that this process has already begun. See also Holder v. Hall, ___ U.S. ___, ___, 114 S.Ct. 2581, 2598, 129 L.Ed.2d 687 (1994) (arguing that the states' racial gerrymandering in response to judicial decisions and DOJ involvement "now promises to embroil the courts in a lengthy process of attempting to undo, or at least minimize, the damage wrought by the system we've created").

Race also need not be the sole motivation behind a redistricting plan before it is subject to further review. This is clear from a reading of Arlington Heights. 429 U.S. at 266, 97 S.Ct. at 563-64. A Shaw claim is an Equal Protection claim; if the Supreme Court were drastically narrowing the Fourteenth Amendment intent requirement articulated in Washington v. Davis and subsequent decisions — specifically for a subset of voting rights cases — it would have said so. More importantly, the standard would be nearly impossible to meet: if race is admittedly one of many valid districting concerns, Shaw, ___ U.S. at ___, 113 S.Ct. at 2826, then plaintiffs could never show that it was the "sole" concern behind a districting plan. A state legislature could always trot out some other traditional districting principle minimally served by the shape in question. Cf. Hays I, 839 F. Supp. at 1202.

The path most useful for negotiating the intricacies surrounding Shaw v. Reno and the Voting Rights Act is the middle one. Both the Supreme Court and Congress have already admitted that the Constitution is not genuinely "color-blind," but it does severely limit the race-based lengths to which a state may go. The document does not suffer racial classifications gladly.

B. Determining if Race was the Overriding Consideration in Creation of the Eleventh District 1. Methods of Proof: Indirect and Direct Evidence

The Supreme Court's use of a district's shape as circumstantial evidence of legislative intent implies that proof might be made by other means. The Shaw v. Hunt court arrived at the same conclusion, finding that proof can be made "by any means, including state concession, bizarre shape, or some combination of the various factors typically used to prove the "intent" element of an Equal Protection claim under Arlington Heights [.]" Shaw v. Hunt, supra, 861 F. Supp. at 431. The Western District of Louisiana concurs: "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." Hays I, 839 F. Supp. at 1195 (footnote omitted). See also Arlington Heights, 429 U.S. at 266, 97 S.Ct. at 564 ("Determining whether invidious discriminatory purpose was a motivating factor demands a sensitive inquiry into such circumstantial and direct evidence of intent as may be available.").

This is clearly the correct approach. Defendants argued early in these proceedings that evidence of the legislature's intent to gerrymander must be inferred from the shape of the Eleventh District itself, and not from direct testimony of those involved with the process. This view finds little support in Shaw v. Reno. The purpose of scrutinizing a district's shape is to glean the intent of the legislature by working backwards: if the district appears uninfluenced by accepted districting principles, as evidenced by its shape, then it must have been influenced by unaccepted ones. The Supreme Court explicitly approved this inferential approach because legislative intent is notoriously difficult — if not logically impossible²³ — to ascertain, and in redistricting cases, the district itself may provide the only firm evidence, albeit circumstantial, of that intent. What the Supreme Court did not do is imbue geography with constitutional significance; the requirement for a successful Equal Protection claim is still intent, however proved. Foreclosing production of direct evidence of intent until Plaintiffs convince the Court that a district looks so weird that race must have dominated its creation is not what Shaw intended. That approach would make district shape a (previously unheard of) threshold to constitutional claims.²⁴

23 See Rogers v. Lodge, 458 U.S. 613, 631-53, 102 S.Ct. 3272, 3282-94, 73 L.Ed.2d 1012 (1981) (Stevens, J., dissenting). Justice Stevens has succinctly stated the problem: "A law conscripting clerics should not be invalidated because an atheist voted for it." Washington v. Davis, 426 U.S. at 253, 96 S.Ct. at 2054 (concurring opinion).

24 Note that a bizarre district shape is indeed a "threshold" for purposes of standing. Without the "harm" caused by an obvious gerrymander, citizens have no basis for suit. See supra, section II.A.1.

2. The Eleventh Congressional District

This stage of our analysis is the easiest to resolve. The amount of evidence of the General Assembly's intent to racially gerrymander the Eleventh District is overwhelming, and practically stipulated by the parties involved.

a. Indirect Evidence

The current configuration of the Eleventh District appears in Joint Exhibit 1. For our purposes, the most significant elements of the Eleventh are the "land bridges" to Dekalb, Richmond and Chatham Counties. On a purely visual level, it is exceedingly obvious to us, and more importantly, to the public, that extraordinary efforts were made to reach those areas for a specific purpose; narrow appendages reach out from the core of the Eleventh to grasp certain goals — the cities of Atlanta, Augusta, and especially Savannah.

Land bridges are not the only bridges which support the Eleventh District's march. The Eleventh soars 18 stories over the water upon the spine of the suspension bridge at Savannah to spill itself over uninhabited Hutchinson Island (which many travelers think is in South Carolina). From Hutchinson the district goes northwest across the marshes, a wildlife refuge, and the river again (this time without a bridge) to a mere point of contiguity with Effingham County. By this means most of the Port facilities are left within the First (historically coastal) District.

The surrounding evidence implies that the aforementioned purpose behind all these bridges, sometimes the only purpose, was race. From intact Screven County, the Eleventh suddenly narrows to a bridge at points barely contiguous and averaging less than two miles in width, until it reaches the City of Savannah, where it branches out to encompass particular sections of that community. See Joint Exh. 1; DOJ Exh. 10, 16-17. Effingham County is actually a growing area, but the part of it within the Eleventh District is largely swamp.²⁵ In most of that strip, the population is literally "zero." Tr. I, at 107 (testimony of Linda Meggers). In Savannah itself, the sections within the Eleventh are heavily populated — with blacks. DOJ.Exh. 17. Chatham County is 38% black; that part of it within the Eleventh is 84% black. Majority white areas of Savannah are circumvented by the district boundaries.²⁶ Id.; Tr. I, at 115-16.

25 The Brooks/McKinney max-black plan employed a wider bridge through Effingham County. During the legislative process, the representative from that area succeeded in narrowing the bridge, to minimize the damage caused in splitting Effingham County by keeping as much of the resident population within the adjacent First Congressional District as possible.

26 In eliminating some of the county splits present in the max-black plan, Ms. Meggers was forced to create even more irregular boundary lines in urban areas, in an effort to include black populations to offset increased white populations in the repaired counties. Tr. I, at 97.

The Court has not been presented with an adequate non-racial rationale for this spectacle. The extension to Savannah is not compact by any credible definition of that term,²⁷ it was not necessitated by natural geographic boundaries, there was no incumbent served by it, there was no economic or political interest thereby enhanced, it split two counties, and it was not necessary to comply with the constitutional one person/one vote requirement. It was urged that many of the district lines follow precinct lines, thus providing a "political," non-racial, explanation for their configuration. While the boundaries of the Eleventh do indeed follow many precinct lines, this is because Ms. Meggers designed the Eleventh District along racial lines, and race data was most accessible to her at the precinct level. Tr. II, at 271-72.

27 We further explore the concept of "compactness" in our discussion of narrow tailoring, section II.C.3, infra.

The Court was also plied with testimony about "communities of interest" involving the black populations of Savannah and other majority-black areas, e.g., Tr. VI, at 26 et seq. (testimony of Rev. Mitchell), 137 et seq. (testimony of Dr. Marsha Darling), that allegedly provide a nonracial justification for the district's current configuration. Here the Defendants and Intervenors faced a Sisyphean task. Initially, the desire to bolster amorphous "communities of interest" could not be sufficient justification for the drafting gymnastics involved in bringing the Eleventh District to Savannah. Second, we remain unconvinced that the poor black populations of coastal Chatham County feel a significant bond to the black neighborhoods of metro Atlanta. Third, even if bonds between these groups exist at certain levels, those links were certainly not the overriding motivation behind construction of a land bridge to Chatham County.

Finally, and most importantly, the Defendants and Intervenors spent much time outlining the racial community of interest shared by black citizens in Georgia. The problem with this tack is that, while partially convincing, such a community of interest is barred from constitutional recognition. To urge this racial identification as a justification for the shape of the Eleventh District is tantamount to simply admitting that race was the overriding consideration in its creation. We have no doubt that black citizens share concerns related to their condition as blacks, e.g., the unusually high crime rate in black communities or combatting racism. Rev. Mitchell presented evidence of religious networking among black congregations. A voting district, however, that is configured to cater to these "black" concerns is simply a race-based voting district. It is based on superficial, racially founded generalizations about what matters to black Georgians. That is, it trafficks in racial stereotypes. We find it ironic and troubling that the state and federal government should expend such effort to convince the Court "that members of the same racial group — regardless of their age, education, economic status, or the community in which they live — think alike, share the same political interests, and will prefer the same candidates at the polls." Shaw, ___ U.S. at ___, 113 S.Ct. at 2827.

Richmond County also shows signs of gerrymandering. From Burke County, the Eleventh District suddenly narrows to a corridor, following the Central Georgia Railroad into the City of Augusta. The corridor, composed of wetlands and some industrial areas, is virtually unpopulated. DOJ Exh. 14; Tr. I, at 100. Upon reaching Augusta, the Eleventh cuts through the city itself, splitting precincts and roping in the heaviest concentrations of black citizens. Richmond County is 42% black; that component inside the Eleventh District is 66% black.

DeKalb County, containing the largest single concentration of blacks in Georgia, suffers a similar fate. From Butts County, the Eleventh gradually narrows to a point less than one half mile wide, and then broadens to cover all of Atlanta's black populations not "needed" for the Fifth District. Tr. I, at 119; DOJ Exh. 8. DeKalb County is only 42.2% black; its constituency within the Eleventh District is 74.6% black.

The narrow section of the district, just south of DeKalb County, provides a clear example of a point we noted earlier. Ms. Meggers, forced to include the black populations of DeKalb County in order to reach the correct black voting age population in the Eleventh District, narrowed the Eleventh to a half mile along the way in order to follow precinct lines. She did this because it was possible, and splitting of precinct lines is generally avoided where possible. Tr. I, at 122-23. Yes, precinct lines provide a nonracial explanation for a district boundary, but, as Ms. Meggers stated, she was "just lucky"; the overriding reason for cutting through that area in the first place was to scoop up black voters. Id. at 82. The fact that Ms. Meggers minimized the damage does not undermine that central motivation.²⁸

28 Defendants also asserted that, since all districts must comply with the one person/one vote requirement articulated in Reynolds v. Sims, 377 U.S. 533, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964), race could not be the dominant or sole motivation behind a redistricting effort. Our response to this argument is that the one person/one vote requirement is a "threshold" requirement, a first-order requirement that must be addressed before considering second-order factors such as race or any traditional redistricting principles. The one person/one vote criterion is inflexible; it simply must be met by every redistricting plan. In contrast, the prominence of secondary factors will vary with the legislative tides. It is among those factors that we debate the relative importance of race as a motivation. One person/one vote compliance is assumed.

Other observations shed light on the racial manipulations behind the Eleventh, most notably the simple one that the total black population of Georgia is 26.96%, while within the Eleventh it is 64.07%. We also note that the vast majority — nearly 80% — of the district's minority population comes from carefully divided counties on its distant fringes. In addition, the Eleventh District avoids the heavy black populations of Macon, just 80 miles from DeKalb, while going to great lengths to include those of Savannah, 260 miles away; the black voters of Macon are necessary to make the Second Congressional District majority-black.

Finally, if not for reasons of race, why split the counties of DeKalb, Wilkes, Richmond, Effingham, Chatham, Twiggs, Baldwin, and Henry? Why split 26 counties at all? With the exception of Texas, Georgia has more counties than any other state in the Union; one would think that such a proliferation would provide ample building blocks for acceptable voting districts without chopping any of those blocks in half. Avoiding such rough surgery did not seem such a herculean feat in any redistricting round before 1992. And why excise Savannah — and just portions of Savannah — from its traditional economic place in the "coastal" region of Georgia? The Eleventh District could have been easily revised to include one or more whole counties while still meeting the one person/one vote requirement, and other traditional districting concerns would have been better served.

We find the above survey indicative of an overriding objective to include minority populations in the Eleventh. In fact, with regard to the extensions to Savannah and Augusta, we find that race was clearly the sole objective behind its creation. At a glance, the appendages of the Eleventh are obviously designed to do something; after cursory exploration, it rapidly becomes clear that the "something" is maximization of black voting strength. As Mr. Armstrong himself admitted, all one must do is

take a . . . map of the State of Georgia shaded for race, shaded by minority concentration, and overlay the districts that were drawn by the State of Georgia and see how well those lines adequately reflect black voting strength.

Tr. IV, at 135. We could not have phrased it better.

b. Direct Evidence

Any attempt to explain the Eleventh District as anything but a far-flung search for black voters utterly collapses under the weight of direct evidence. One need only review DOJ's objection letters and our findings of fact to see that DOJ spent months demanding purely race-based revisions to Georgia's redistricting plans, and that Georgia spent months attempting to comply. The testimony at trial reflects this sorry process, and it is compelling. We need review but a few examples from the record.

The State of Georgia simply confessed to its racial motivations. In response to Plaintiffs' Request, State made the following admissions of fact:

The State would not have added those portions of Effingham and Chatham Counties that are now in the present Eleventh Congressional District but for the need to include additional black population in that district to offset the loss of black population caused by the shift of predominantly black portions of Bibb County in the Second Congressional District which occurred in response to the Department of Justice's March 20th, 1992, objection letter.

Tr. I, at 11-12. And

To the extent that precincts in the Eleventh Congressional District are split, a substantial reason for their being split was the objective of increasing the black population of that district.

Id. at 12.

Speaker Thomas Murphy also admitted that the Assembly built its third districting plan along racial lines.

What we did is we went into counties and precincts and picked up pockets of African-Americans to make a strong district with voting age black population so that it would guarantee a black would be elected from there.

Tr. II, at 62. Speaker Murphy stated that the boundaries of Chatham, Richmond, DeKalb, Baldwin, Twiggs, and Wilkes Counties were "all drawn based upon race." Id. at 72. He was not alone in his frankness. Every legislator testifying in these proceedings admitted that the objectives behind the districting plan were racial ones. Max-black supporters presumably agreed, including Rep. Cynthia McKinney and various members of the Georgia Black Caucus. Linda Meggers, not an elected official but the architect behind the districts in question, readily stated on numerous occasions that she designed the Eleventh District to root out black populations and avoid white ones. See Tr. I, at 106-07 (Chatham County), 101-02 (Richmond County), 124-25 (Baldwin County), 270-71 (DeKalb County), 272-73 (on sole purpose of land bridges as connections to black populations). See also id. at 74 (Second Congressional District), 82 (same). Ms. Meggers routinely found the "blackest" precincts first, and then worked backwards by racial percentage. Id.

Rep. Sonny Dixon, who "oversaw the extension of the Eleventh District into Chatham County," Tr. IV, at 159, gave powerful and credible testimony on that subject, and was quite clear as to how the current Eleventh District came to be:

MR. CHESIN [for Plaintiffs]: What was the reason for this extension [into Chatham County]?

REP. DIXON: Purely to, as the Justice Department objection letter stated, to connect the second largest concentration of blacks with the rest of the Eleventh District.

MR. CHESIN: Was there anything else besides race?

REP. DIXON: None whatsoever.

Id. at 159-60. He detailed the extreme efforts made to stretch the Eleventh to Savannah without absorbing any white voters along the way, id. at 161-64, and the literally block by block search for black voters to add to the Eleventh District.

Rep. Dixon's testimony presents a picture which, under virtually any other set of circumstances, would be clearly impermissible, almost unthinkable. The concept of government allocations on the basis of race, coupled with drawing lines tracing concentrations of black citizens, smacks of government-enforced ghettoization. The contours of the Eleventh District reviewed above are so dramatically irregular as to permit no other conclusion than that they were manipulated along racial lines. This conclusion is cemented with the copious amounts of direct evidence discussed here. Consequently, we must now determine whether this kind of racial classification is narrowly tailored to comply with some compelling state interest.

C. Constitutional Analysis

Race-based laws are subject to strict scrutiny even if "remedial" in nature. Richmond v. J.A. Croson Co., 488 U.S. 469, 109 S.Ct. 706, 102 L.Ed.2d 854 (1989) (plurality opinion); Wygant v. Jackson Board of Education, 476 U.S. 267, 273, 106 S.Ct. 1842, 1846-47, 90 L.Ed.2d 260 (1985) (plurality opinion); United Jewish Organizations, 430 U.S. at 172, 97 S.Ct. at 1013 (Brennan, J., concurring in part). As stated in Croson, "[a]bsent searching judicial inquiry . . ., there is simply no way of determining what classifications are 'benign' or 'remedial' and what classifications are in fact motivated by illegitimate notions of racial inferiority or simple racial politics." Id. 488 U.S. at 493-94, 109 S.Ct. at 721. Once revealed as a racial classification, a law can survive constitutional review only upon "extraordinary justification," that is, a showing that the law is both necessary and narrowly tailored to further a compelling state interest. See Shaw v. Reno, ___ U.S. at ___, 113 S.Ct. at 2832; Palmore v. Sidoti, 466 U.S. 429, 432, 104 S.Ct. 1879, 1881-82, 80 L.Ed.2d 421 (1984); Fullilove v. Klutznick, 448 U.S. 448, 491, 100 S.Ct. 2758, 2781, 65 L.Ed.2d 902 (1980).

The State Defendants seem to propose only the goal of proportional representation as their compelling justification for the current Eleventh Congressional District. Def. Prop. Findings, ¶ 122. The Intervenors, however, suggest VRA compliance and eradicating the effects of past or present racial discrimination as interests sufficiently compelling to justify Georgia's actions in this case. It is clear to us that the only interest the General Assembly had in mind when drafting the current congressional plan was satisfying DOJ's preclearance requirements. The articulated "compelling" justifications appear to be post hoc rationalizations. Nevertheless, we will address all possibilities.

1. A Preliminary Issue: Burden of Proof

The burden of proof as to whether race was the overriding consideration behind the current districting plan rested squarely with the Plaintiffs. At the strict scrutiny stage, we agree with the court in Shaw v. Hunt that, while burdens of production shift to the Defendants, the ultimate burden of persuasion remains with Plaintiffs:

In . . . a reverse-discrimination case, as in any other Equal Protection case, "[t]he ultimate burden remains with the [plaintiff] to demonstrate the unconstitutionality of [the] affirmative action program." Wygant, 476 U.S. at 277-78 [ 106 S.Ct. at 1849] . . . Proof that the challenged law or policy is race-based gives rise to a presumption that it is unconstitutional and shifts to the state the burden of "demonstrating" that its use of race was justified by a compelling governmental interest. Croson, 488 U.S. at 505 [ 109 S.Ct. at 727-28] (majority). But the burden thus shifted is one of production only, not persuasion[.]

Shaw v. Hunt, supra, 861 F. Supp. at 436. Plaintiffs still bear the burden of persuading the court that the State's evidence did not support a finding that the redistricting plan was narrowly tailored to a compelling state interest.

2. Potential Compelling State Interests

a. Proportional Representation

We only briefly address this argument, since (1) we encounter it later in our discussion of narrow tailoring, and (2) it clearly does not provide a state interest sufficient to pass constitutional muster.

Proportional representation simply is not a constitutional requirement. See, e.g., Mobile v. Bolden, 446 U.S. at 76, 100 S.Ct. at 1504 ("The Equal Protection Clause of the Fourteenth Amendment does not require proportional representation as an imperative of political organization."); Whitcomb v. Chavis, 403 U.S. 124, 149, 91 S.Ct. 1858, 1872, 29 L.Ed.2d 363 (1971); Turner v. Arkansas, 784 F. Supp. 553, 577 (E.D.Ark. 1991) (listing cases). Nor is it a statutory requirement. See 42 U.S.C. § 1973b (prohibiting vote dilution, and stating "nothing in this section establishes a right to have members of a protected class elected in numbers equal to their proportion in the population"). To erect the goal of proportional representation is to erect an implicit quota for black voters. Far from a compelling state interest, such an effort is unconstitutional. See, e.g., Regents v. Bakke, 438 U.S. at 307, 98 S.Ct. at 2757.

The State believes in a compelling state interest in achieving proportionality apart from any requirement of the voting Rights Act. State Def. Findings, ¶ 122. In response to an inquiry as to what compelled the creation of racially proportional voting districts, the State's tautological answer is to proffer a desire to draft racially proportional voting districts, and insist that "the Constitution does not prohibit a state from enacting such a plan." True, but neither does the Constitution — nor the VRA — require it. This distinction is stressed in Shaw v. Reno, ___ U.S. at ___, 113 S.Ct. at 2830, and is partially the lesson of Johnson v. De Grandy, ___ U.S. ___, 114 S.Ct. 2647, 129 L.Ed.2d 775 (1994), on which both the State and DOJ rely. De Grandy acknowledges proportionality as a prima facie indication of nondilution under section 2, and as something to consider within the "totality of the circumstances" surrounding a vote dilution claim. Lack of proportionality may imply dilution, but that inference is far from an independent and compelling state interest in racially gerrymandering voting districts.

b. Eradicating the Effects of Past Discrimination

With the rise of race-based programs intended to remedy the effects of past discrimination, it became necessary for the Supreme Court to decide when a history of "societal discrimination" provided a sufficiently compelling state interest for the law in question. A generalized claim of societal discrimination cannot justify a racial classification. Wygant, 476 U.S. at 276, 106 S.Ct. at 1848. There must be some "particularized findings" of past discrimination, providing a "strong basis in evidence" for the state's conclusion that remedial action is necessary. Id.; Croson, 488 U.S. at 500, 504, 109 S.Ct. at 725, 727 (majority opinion).

Such a finding might be possible in our case; we have given formal judicial notice of the State's past discrimination in voting, and have acknowledged it in the recent cases. See Brooks v. State Bd. of Elections, 848 F. Supp. 1548, 1560-61, 1571 (S.D.Ga. 1994). See also Rogers v. Lodge, 458 U.S. 613, 102 S.Ct. 3272, 73 L.Ed.2d 1012 (1981). The problem is that (1) the General Assembly never articulated such lofty goals during the 1990-92 redistricting, having made clear that it would not have enacted the current plan but for DOJ demands, and (2) we conclude that a compelling state interest in remedying prior discriminatory voting practices does not exist independent of the Voting Rights Act.

The Supreme Court declined to fully explore this latter issue in Shaw v. Reno, but did take time to "note" that

only three Justices in UJO were prepared to say that States have a significant interest in minimizing the consequences of racial block voting apart from the requirements of the Voting Rights Act. And those three Justices specifically concluded that race-based districting, as a response to racially polarized voting, is constitutionally permissible only when the State "employ[s] sound districting principles," and only when the affected racial group's "residential patterns afford the opportunity of creating districts in which they will be in the majority."

Shaw, ___ U.S. at ___, 113 S.Ct. at 2832 (quoting UJO, 430 U.S. at 167-68, 97 S.Ct. at 1011) (opinion of White, J., joined by Stevens and Rehnquist, JJ.). Justice Brennan concurred in UJO:

[I]f and when a decisionmaker embarks on a policy of benign racial sorting, he must weigh the concerns that I have discussed against the need for effective social policies promoting racial justice in a society beset by deep-rooted racial inequities. But I believe that Congress here adequately struck that balance in enacting the carefully conceived remedial scheme embodied in the Voting Rights Act.

Id. at 175, 94 S.Ct. at 1014-15. We take our cue from both the Shaw v. Reno majority's caveat and Justice Brennan's faith in the Voting Rights Act as a justification for race-based legislative preferences. The VRA is an expansive remedial scheme imposing federal authority over much of the country's state and local voting systems. In effect, it formalizes, codifies, and universally imposes a "compelling state interest" to redress historically persistent discriminatory voting practices. See, e.g., South Carolina v. Katzenbach, 383 U.S. 301, 310-15, 86 S.Ct. 803, 809-12, 15 L.Ed.2d 769 (1965) ("The Voting Rights Act of 1965 reflects Congress' firm intention to rid the country of racial discrimination in voting."). Under the guise of section 5, it even directs that "state interest" to those regions where it is most "compelling."

Any independent state interest in the remedial revision of voting laws is subsumed within that broad federal legislation. If a state's remedial program meets the requirements of the VRA, an independent compelling justification is redundant; if the program exceeds the requirements of the VRA, the separate interest in further increasing minority voting strength is no longer compelling. The VRA places a necessary leash on race-based remedial practices in the voting sphere; the Supreme Court has acknowledged that remedial legislation can cause more harm than good, and requires such laws to be narrowly tailored and clearly linked to specific instances of past discrimination. See Wygant, 476 U.S. at 274-76, 106 S.Ct. at 1847-48 (requiring a "strong basis in evidence" for any proffered state interest in rectifying prior discrimination); Bakke, 438 U.S. at 298, 98 S.Ct. at 2752 (observing that "preferential programs may only reinforce common stereotypes holding that certain groups are unable to achieve success without special protection based on a factor bearing no relationship to individual worth."). Considering the unparalleled significance of the right to vote, this implicit regulation is beneficial to the nation. The VRA ensures that states honor the right to vote; our recognition of no other compelling justification for race-based voting remedies ensures that states do not go too far.²⁹

29 In contrast to our view, the majority in Shaw v. Hunt found recognition of a compelling state interest beyond the VRA necessary because the VRA might not reach all instances in which race-based redress is needed. They presented the example of a state that

has a history of official racial discrimination in its electoral system, which has resulted in the virtual exclusion of members of a particular racial minority from participation in its political processes, but it knows that the creation of majority-minority districts is not required by the [nonretrogression element] of section 5, because it has never had such districts before, and that the relevant minority group cannot show that section 2 requires the creation of any majority-minority districts, because it is too widely dispersed to constitute a majority in a single-member district that is "geographically compact" under Gingles.

Shaw v. Hunt, supra, 861 F. Supp. at 444. Our response is that this hypothetical state rightly has no compelling interest in gathering together its widely dispersed minority population. To do so, it would have to traffick in a superficial stereotype impugned by Shaw v. Reno — that widely dispersed black populations will vote alike simply because they are black. The vote dilution provisions of the VRA implicitly recognize this point by not requiring a different result. Minority populations must be "geographically compact" before a successful section 2 claim can be made, Thornburg v. Gingles, 478 U.S. 30, 50-51, 106 S.Ct. 2752, 2766-67, 92 L.Ed.2d 25 (1986), and this requirement is designed precisely to bar such claims where the minority population is "substantially integrated" or "spread evenly" throughout the challenged district. See id. at 50 and n. 17, 106 S.Ct. at 2766 and n. 17. Finally, we are here presented with a textbook example of "hard cases" making "bad law," Northern Securities Co. v. United States, 193 U.S. 197, 363, 24 S.Ct. 436, 467, 48 L.Ed. 679 (1904) (Holmes, J., dissenting); the improbability of the above scenario indicates its dubious value as scaffolding for constitutional arguments.

c. VRA Compliance

The Shaw v. Reno majority stops short of declaring that compliance with the Voting Rights Act is, a fortiori, a compelling state interest. The Court allows that "[t]he States certainly have a very strong interest in complying with federal antidiscrimination laws that are constitutionally valid as interpreted and applied." Shaw, ___ U.S. at ___, 113 S.Ct. at 2830. Fully steeped in the argot of constitutional law, we do not doubt that the Court intentionally avoided the word "compelling," instead opting for the more cautious "very strong." This suggests to us that VRA compliance might be a compelling state interest under some circumstances, but that there are certain necessary predicates; the Court warns that a redistricting plan satisfying section 5 may still be found unconstitutional, and that legislatures are not thereby granted "carte blanche to engage in racial gerrymandering[.]" Id. at ___, 113 S.Ct. at 2831 (citing the VRA itself and Supreme Court precedents). Beyond these cautionary observations, no explicit guidance is given as to when VRA compliance constitutes a compelling state interest.

A most useful hint, however, comes from Justice O'Connor's injunction that courts "bear in mind the difference between what the law permits, and what it requires." Id. at ___, 113 S.Ct. at 2830. Essentially, the point here is that the VRA cannot justify all actions taken in its name. It will present a compelling justification for certain race-based remedial measures, but not for all that its provisions might potentially "permit." The court in Shaw v. Hunt translates this aspect of Shaw v. Reno into a useful standard:

We think it clear that a state has a "compelling" interest in engaging in race-based redistricting to give effect to minority voting strength whenever it has a "strong basis in evidence" for concluding that such action is "necessary" to prevent its electoral districting scheme from violating the Voting Rights Act.

Shaw v. Hunt, supra, 861 F. Supp. at 437. See Hays I, at 1217 (concurring opinion). That court arrives at this interpretation by marrying the standard for determining if past discrimination provides a compelling state interest with recent Supreme Court pronouncements on remedial race-based measures. See Wygant v. Jackson Board of Education, 476 U.S. 267, 276, 106 S.Ct. 1842, 1848, 90 L.Ed.2d 260 (1986) (requiring "particularized findings" of societal discrimination before concluding that a compelling state interest exists for remedying it); Croson, 488 U.S. at 491, 109 S.Ct. at 720 (considering remedial actions in contracting); Bakke, 438 U.S. at 307, 98 S.Ct. at 2757 (considering educational affirmative action programs). This standard makes sense, but is indicative of a puzzling component of Shaw v. Reno explored below.

The Effect of Shaw v. Reno on the Strict Scrutiny Standard in the Context of VRA Compliance

If we assume that the VRA is constitutional,³⁰ and that it does, as Shaw strongly implies, provide sufficiently compelling justification for race-based redistricting in some contexts, then the narrow tailoring stage of the constitutional analysis logically overshadows the compelling interest stage as the vital point of contention. The Shaw v. Hunt standard reflects this shift: it is unlikely, especially considering DOJ's current standards for preclearance under section 5, that a state could not provide solid evidentiary support for concluding that its redistricting efforts were "necessary." As the North Carolina court argues, rejection of a redistricting proposal by DOJ or the D.C. District Court would constitute "a strong basis in evidence" justifying increasingly race-based redistricting. Indeed, such measures would then appear "necessary" for VRA compliance, since preclearance would not otherwise be forthcoming.

30 No risky assumption. See, e.g., Chisom v. Roemer, 501 U.S. 380, 111 S.Ct. 2354, 115 L.Ed.2d 348 (1991) (upholding constitutionality of "purpose" requirements of VRA sections 2 and 5); South Carolina v. Katzenbach, 383 U.S. 301, 334-35, 86 S.Ct. 803, 821-22, 15 L.Ed.2d 769 (1966) (holding that Congress did not abuse its power in creating, preclearance requirement under VRA section 5, since it was a "legitimate measure" taken under "compulsion of . . . unique circumstances").

A "prima facie" compelling state interest would thereby be created, "prima facie" because the race-based measures have not been verified as necessary; it is only the State defendant — and DOJ — that assert as much. To determine if such was indeed the case, a court would have to pursue a narrow tailoring inquiry, and decide whether the districts, as precleared and enacted, were genuinely "reasonably necessary," Shaw, ___ U.S. at ___, 113 S.Ct. at 2831, that is, narrowly tailored, to comply with the VRA. If they were not found actually necessary, then the State — and DOJ — were wrong in deciding that they were, and suddenly no compelling interest existed; if the measures were found necessary, then both DOJ's interpretation of the VRA and the State's adherence to it had been correct, and so a compelling interest had existed.³¹

31 We disagree with Chief Judge Voorhees that this approach results in a "[b]lind deference to the administrative findings of the United States Attorney General" that confers immunity from "constitutional scrutiny[.]" Shaw v. Hunt, supra, dissenting opinion, 861 F. Supp. at 487. First, DOJ was expressly empowered by Congress to render section 5 decisions on par with those of the D.C. district court. 28 C.F.R. § 51.52. Labeling them as "administrative findings," Bakke, 438 U.S. at 305, 98 S.Ct. at 2756 (opinion of Powell, J.), does not diminish their potency, and the Supreme Court does not seem to underestimate their importance. See, e.g., UJO, 430 U.S. at 175, 97 S.Ct. at 1014-15 (Brennan, J., concurring in part). Second, it is obvious that, since DOJ enforces the VRA via its preclearance duties, its demands will dictate what it is "necessary" for the State to do under the Act. Again, Congress realized this when the VRA was enacted. Third, our analysis in the text above demonstrates that no constitutional "immunity" inheres from the standard articulated by the majority in Shaw v. Hunt. It does somewhat alter the balance of factors in the strict scrutiny analysis, but the State is granted no added protection thereby. Close examination of potentially compelling interests is simply held in abeyance pending a determination of narrow tailoring. Finally, the Attorney General is not hereby granted "de facto ability to determine the constitutional scope of federal legislation" anymore than she is granted that ability under any other federal law that DOJ is entrusted to enforce.
Shaw v. Reno allows that the VRA is potentially a compelling interest, but only if race-based changes made in its name are "reasonably necessary" to comply with it. Yes, the State can — and should — independently assess whether its plans reasonably satisfy the Act. The State did so in the instant case, and the result was predictable: DOJ's assessment trumped that of the General Assembly, since without DOJ's acquiescence, there are no congressional elections in Georgia. There was no "blind deference" here. Georgia was a reluctant participant.

The end result here is that in scenarios involving jurisdictions subject to section 5, a compelling interest is initially assumed, since the plans in question could not have been enacted without VRA "compliance" as interpreted by the Justice Department. We realize that the Shaw v. Reno majority was careful to say that "States . . . have very strong interest in complying with federal antidiscrimination laws that are constitutionally valid as interpreted and applied," Shaw, ___ U.S. at ___, 113 S.Ct. at 2830 (emphasis added), and that DOJ's "maximization" interpretation in this case was potentially unconstitutional.³² The interpretation was improper, however, because it compelled legislative efforts not reasonably necessary/narrowly tailored to the written dictates of the VRA and attendant regulations, not because it required race-conscious redistricting. We are constantly returned to that crucial tailoring aspect of the strict scrutiny rubric, and our temporary compelling interest assumption is merely a stepping stone to get us there. Cf. Hays I, 839 F. Supp. at 1206 (assuming presence of a compelling state interest in order to reach the narrow tailoring stage).

32 We realize that decisions of the Attorney General are not reviewable by this Court. See Morris v. Gressette, 432 U.S. 491, 97 S.Ct. 2411, 53 L.Ed.2d 506 (1977).

It is not seriously debated that the General Assembly designed the current congressional districts specifically to comply with DOJ's preclearance requirements. Reviewing our factual findings, it is evident that Georgia, wrangling with the DOJ Voting Section for over a year, did what was minimally "necessary" to secure preclearance, and putatively, to comply with the VRA. Consequently, there was a putatively compelling interest behind the drafting of the current Eleventh Congressional District, pending a final determination of whether DOJ's demands comported with the VRA.³³ We now leave this stage and turn to whether Georgia's 1992 congressional redistricting plan was actually required by the VRA and Supreme Court precedent; the Plan will live or die on the results of our narrow tailoring inquiry.

33 "Reliance on possibly invalid applications of the Voting Rights Act by the Department of Justice cannot create a [genuinely] compelling state interest. If so, the Department of Justice and various States could sidestep the holdings of Croson, Gingles, and Shaw with ease." Hays I, 839 F. Supp. at 1217.

3. Narrow Tailoring

A variety of factors are generally relevant to whether a state action is narrowly tailored to further a compelling state interest,³⁴ but in the voting context, the factors are somewhat altered by the presence of the VRA. Our inquiry here will focus on two issues: (1) whether the current plan contains more majority black districts than reasonably necessary to comply with the VRA, and (2) whether the existing majority black districts contain larger concentrations of minority voters than reasonably necessary to give those voters a realistic opportunity to elect candidates of their choice. See Shaw v. Hunt, supra, 861 F. Supp. at 445-46; Hays I, 839 F. Supp. at 1206-08.

34 These factors include (1) the necessity of the measure, (2) the efficacy of alternative, race-neutral measures, (3) the availability of more narrowly tailored options, (4) the flexibility and duration of the measure, and (5) the impact of the measure on the rights of third parties. Hays, 839 F. Supp. at 1208, 1215. See Croson, 488 U.S. at 510-11, 109 S.Ct. at 730-31; United States v. Paradise, 480 U.S. 149, 171, 107 S.Ct. 1053, 1066-67, 94 L.Ed.2d 203 (1987).

a. The Significance of Maximization

A congressional districting plan is not narrowly tailored if it affects the rights and interests of citizens more than "reasonably necessary" to further the compelling state interest advanced by the state. See Shaw, ___ U.S. at ___, 113 S.Ct. at 2831 (setting out "reasonably necessary" language); Hays I, 839 F. Supp. at 1208 (describing this element as the "essence of narrow tailoring in the redistricting context"). Shaw draws a comparison with UJO, where four justices decided that New York's creation of additional majority-minority districts was constitutional, because the plaintiffs had failed to show that the state "did more than the Attorney General was authorized to require it to do[.]" 430 U.S. at 162-63, 97 S.Ct. at 1008-09. In other words, plaintiffs failed to show that the State's plan did more than was necessary to satisfy the compelling interest in complying with the VRA. That showing is much easier here, where DOJ stretched the VRA farther than intended by Congress or allowed by the Constitution, and had the General Assembly firmly in tow.

On first appraisal, our factual finding that the Georgia plan reflects a DOJ maximization agenda³⁵ bodes ill for any arguments that the Eleventh District is narrowly tailored to comply with VRA requirements. Consider the regulations promulgated by Congress to assist DOJ in properly enforcing the Act. Among the numerous racial factors specific to redistricting, DOJ must also consider

35 We are by no means the first Court to criticize DOJ's maximization propensities. See Shaw v. Hunt, supra, Dissenting Opinion; Hays I, 839 F. Supp. at 1196-97 n. 21; Turner v. Arkansas, 784 F. Supp. at 561 (quoting Abigail Thernstrom, Washington Post, Sept. 23, 1991, at A11).

(e) The extent to which available alternative plans satisfying the jurisdiction's legitimate governmental interests were considered.

(f) The extent to which the plan departs from objective redistricting criteria set by the submitting jurisdiction, ignores other relevant factors such as compactness and contiguity, or displays a configuration that inexplicably disregards available natural or artificial boundaries.

28 C.F.R. § 51.59. Congress intended these two criteria as indicators of whether a legislature was attempting to gerrymander in favor of white voters, but they are equally useful tools in our context. We will not rehash our factual recitation here, but observe that DOJ clearly disregarded these intended restraints on gerrymandering. DOJ's requirements, based on the max-black proposal, had nothing at all to do with any criterion but race. See Pltf.Exh. 2, 4, 6 (DOJ objection letters). Some fundamental "objective redistricting criteria," like the one person/one vote, single-member, and contiguity requirements, were satisfied along the way, but every factor that could realistically be subordinated to racial tinkering in fact suffered that fate. The district is not compact, it disregards economic boundaries, and it ignores county and precinct lines at will when needed to reach black neighborhoods.

We find it ironic that Congress was apparently worried that the legislature might ignore alternative plans that were able to boost black voting strength while satisfying legitimate nonracial interests. Here, it was DOJ that demanded an alternative plan that boosted black voting strength and completely ignored legitimate nonracial interests. In a similar vein, Congress uses the word "inexplicably," in subsection (f), apparently envisioning a district with lines unexplainable as anything other than an effort to exclude black voters; here we are confronted with a district unexplainable as anything other than an effort to exclude white voters.

These observations alone lead us to conclude that the current congressional plan is not narrowly tailored to what the VRA actually, reasonably, requires. We continue, however, with a more specific look at the gulf between the VRA and the enacted plan.

b. Section 5 of the Voting Rights Act

Section 5 of the VRA prohibits a State or political subdivision subject to section 4 of the Act from enforcing any change in voting practices unless it has obtained a declaratory judgment from the D.C. District Court that such a change "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," or has secured preclearance from the Attorney General under the same standard. While the "purpose" prong, essentially eclipsed by section 2 of the Act,³⁶ requires a showing that the proposed plan was not designed to dilute minority voting strength, the "effect" prong is aimed at barring any redistricting plan that would lead to "a retrogression in the position of racial minorities with respect to their effective exercise of the electoral franchise." 28 C.F.R. § 51.54(a) (making retrogression the standard for finding discriminatory effect under section 5, and citing Beer v. United States, 425 U.S. 130, 141, 96 S.Ct. 1357, 1363-64, 47 L.Ed.2d 629 (1976)). See also McCain v. Lybrand, 465 U.S. 236, 247, 104 S.Ct. 1037, 1044-45, 79 L.Ed.2d 271 (1984) (making purpose/effect distinction); Port Arthur v. United States, 459 U.S. 159, 168, 103 S.Ct. 530, 535-36, 74 L.Ed.2d 334 (1982) (discussing purpose prong).

36 The only difference is that under the section 5 analysis the burden of proof is not on affected minority groups, but on the jurisdiction seeking preclearance. See 28 C.F.R. § 51.52 (placing section 5 burden of proof on "submitting authority").

It is abundantly clear to this Court that Georgia's current redistricting plan exceeds what is reasonably necessary to avoid retrogression under section 5. DOJ readily admitted that neither the first nor the second redistricting plan passed by the Assembly had a retrogressive effect on minority voting strength. Tr. V, at 36-38; Pltf. Exh. 99 (Requests for Admissions from DOJ). Those first two plans provided for two majority black districts, but the third, current plan — in keeping with the max-black proposals — provides three such districts, prompting us to conclude that three majority black districts are not reasonably necessary to comply with the VRA. The Shaw v. Reno majority stated: "A reapportionment plan would not be narrowly tailored to the goal of avoiding retrogression if the State went beyond what was reasonably necessary to avoid retrogression." Shaw v. Reno, ___ U.S. at ___, 113 S.Ct. at 2831. Having created more majority black voting districts than necessary to avoid retrogression, the State of Georgia enacted a congressional districting plan that was not narrowly tailored to the compelling state interest of complying with the VRA. Cf. Beer, 425 U.S. at 141, 96 S.Ct. at 1363-64 ("It is thus apparent that a legislative reapportionment that enhances the position of racial minorities with respect to their effective exercise of the electoral franchise can hardly have the 'effect' of diluting or abridging the right to vote on account of race within the meaning of section 5.").

Further, DOJ's admission is not necessary to make clear the lack of retrogressive effect. Under the previous congressional districting scheme, there was one majority-black district among the ten districts in Georgia, or 10% majority-black districts. In a plan with eleven districts, two majority-black districts would constitute 18.18% of the total — quite an improvement.³⁷ Federal regulations state that "[i]n determining whether a submitted change is retrogressive the Attorney General will normally compare the submitted change to the voting practice or procedure in effect at the time of the submission." 28 C.F.R. § 51.54(b). Apparently, neither DOJ nor the General Assembly used this simple guide to Section 5 compliance.

37 Under the first and second proposals, the Second Congressional District was a black "influence" district augmenting the black voting strength already harnessed in the Fifth and Eleventh.

The current districting plan overstepped the requirements for section 5 compliance because it was designed to secure proportional representation for black voters in Georgia, not adhere to the VRA. Three majority-black districts constitute 27.27% of the total eleven, and blacks constitute 26.96% of the total Georgia population. DOJ used its section 5 preclearance prerogative as a tool for forcing that standard upon the General Assembly, thereby setting an extremely dangerous precedent on at least two counts.

First, as noted earlier, proportional representation is in no way a constitutional or statutory requirement. See De Grandy, ___ U.S. ___, 114 S.Ct. 2647; Mobile, 446 U.S. at 55, 100 S.Ct. at 1493; Beer, 425 U.S. at 136, 96 S.Ct. at 1361. See also 42 U.S.C. § 1973b. The Supreme Court has in fact rejected any proposed requirement of proportional representation or vote maximization throughout the history of the VRA. See, e.g., White v. Regester, 412 U.S. 755, 765-66, 93 S.Ct. 2332, 2339-40, 37 L.Ed.2d 314 (1973); Whitcomb v. Chavis, 403 U.S. 124, 148-56, 91 S.Ct. 1858, 1871-76, 29 L.Ed.2d 363 (1971). And most recently:

Operating under the constraints of a statutory regime in which proportionality has some relevance, States might consider it lawful and proper to act with the explicit goal of creating a proportional number of majority-minority districts in an effort to avoid section 2 litigation. . . . The Department of Justice might require (in effect) the same as a condition of granting preclearance, under section 5 of the Act, 42 U.S.C. § 1973c, to a State's proposed legislative redistricting. Those governmental actions, in my view, tend to entrench the very practices and stereotypes the Equal Protection Clause is set against.

De Grandy, ___ U.S. at ___, 114 S.Ct. at 2666 (Kennedy, J., concurring) (emphasis added). The potential harm of a proportional representation benchmark is not easily overestimated. Such a standard dictates racial quotas for our democratic institutions based on the percentage of each race in the population at large. No, there is no guarantee that majority-minority districts will yield minority representatives, but their conscious construction as entitlements for each racial group causes the societal damage against which we warn today. If efforts to require proportional representation of minorities in democratic institutions are not stopped with clarity and force, they will divide this country into a patchwork of racial provinces, and ensure that elected officials represent races before they represent citizens. See Holder, ___ U.S. at ___ — ___, 114 S.Ct. at 2598-99. The VRA neither intends nor requires the devolution of voting rights into racial bargaining chips to be bickered over by special interests, and we will not support that cause.

Second, while the Shaw v. Hunt majority is confident that race-based redistricting plans "are inherently temporary in nature," Majority Opinion, 861 F. Supp. at 447, we are equally sure that any redistricting plan designed for proportionality, that is upheld by the Court today, will not be easily uprooted in the foreseeable future. We are aware that the decennial census prompts revisions to many districting plans, but the VRA prevents any voting changes that have a dilutive or retrogressive effect upon minority voting strength. Consequently, any plan approved today will become the absolute baseline for subsequent changes through 2007, 42 U.S.C. § 1973b(a)(8); in 2000, if the Georgia population increases and the Georgia Assembly does not add a fourth majority-black district, surely it will be subject to litigation under section 2 and rejection under section 5. Nor do we find it likely that Congress will commit the wildly unpopular act of substantially revising VRA nondilution and nonretrogression provisions in 1997. See id. § 1973b(a)(7). The "bailout" mechanism available to jurisdictions covered by section 5 does impose some eventual limit on what is required, id. § 1973b(a)(1), but even if released from section 5 coverage, Georgia would not then be free to dismantle its proportionally representative districting scheme. Upholding the plan today would create a fairly permanent race-based electoral system.

c. Section 2 of the Voting Rights Act

Having thoroughly denounced the cause of proportional representation, we do realize that it plays an important role in vote dilution claims under section 2 of the VRA. Without it, there is no statistical norm against which to compare the current state of minority voting prowess. De Grandy, supra, ___ U.S. at ___, 114 S.Ct. at 2664-65 (Kennedy, J., concurring) (citing Supreme Court cases). See also Holder, ___ U.S. at ___, 114 S.Ct. at 2597 ("[T]he mathematical principle driving the results in our cases, is undoubtedly direct proportionality."); Mary A. Inman, Change Through Proportional Representation: Resuscitating a Federal Electoral System, 141 U.Pa.L.Rev. 1991, 2050-51 (1993) (arguing that despite the VRA, "the core value underlying Gingles' three preconditions is a right to proportional representation — but only for compact, cohesive, and sizable minority groups." (footnotes omitted)). The De Grandy majority discusses proportional representation at length, finding that while proportionality is probative evidence that minority voters have an "equal opportunity" to participate in political processes and elect their preferred candidates, its significance "may vary with other facts." ___ U.S. at ___, 114 S.Ct. at 2661. While sometimes sufficient to show nondilution, it is not required to achieve it. Id. at ___ — ___, 114 S.Ct. at 2660-62.

Congress enacted section 2 of the VRA to provide teeth to the Fifteenth Amendment, a constitutional injunction historically spurned by state and local governments. See Katzenbach, 383 U.S. at 308, 86 S.Ct. at 808 et seq. (reviewing history). Section 2(a) prohibits the imposition of any electoral practice or procedure that "results in a denial or abridgement of the right of any citizen . . . to vote on account of race or color[.]" 42 U.S.C. § 1973(a). Section 2(b) specifies that section 2(a) is violated if:

[B]ased 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.

Id. Section 2(a) adopts a "results" test, "providing that proof of discriminatory intent is [unnecessary] to establish any violation of the section. Section 2(b) provides guidance about how the results test is to be applied." Chisom v. Roemer, 501 U.S. 380, 395, 111 S.Ct. 2354, 2364, 115 L.Ed.2d 348 (1991); Thornburg v. Gingles, 478 U.S. 30, 35, 106 S.Ct. 2752, 2758, 92 L.Ed.2d 25 (1986). In the context of single-member voting districts, "the usual device for diluting minority voting power is the manipulation of district lines." Voinovich v. Quilter, ___ U.S. ___, ___, 113 S.Ct. 1149, 1155, 122 L.Ed.2d 500, 511 (1993). Minority voting strength might be "cracked," fragmented and dispersed among two or more districts where it constitutes an ineffective minority of the electorate, or "packed," concentrated in one or a few districts where it constitutes an excessive majority, thus "bleaching" the surrounding districts of minority voters. See id.; Gingles, 478 U.S. at 46 n. 11, 106 S.Ct. at 2764 n. 11.

Gingles provides the essential framework for establishing a vote dilution claim against at-large, multimember districts; it has since been extended to single-member districts.³⁸ Growe v. Emison, ___ U.S. ___, ___ — ___, 113 S.Ct. 1075, 1084-85, 122 L.Ed.2d 388 (1993). There are three "threshold conditions": (1) that "[the minority group] is sufficiently large and geographically compact to constitute a majority in a single-member district"; (2) "that [the minority group] is politically cohesive"; and (3) "that the white majority votes sufficiently as a bloc to enable it . . . usually to defeat the minority's preferred candidate." Gingles, 478 U.S. at 50-51, 106 S.Ct. at 2766-67. If the plaintiffs in a vote dilution suit establish these conditions, the court then considers other factors relevant to determining whether, "under the totality of the circumstances," minorities have been denied an "equal opportunity" to "participate in the political process and to elect representatives of their choice." 42 U.S.C. § 1973b. See id., 478 U.S. at 80, 106 S.Ct. at 2781-82. These secondary factors can include a history of discriminatory voting practices, racially polarized voting, the relative presence of minorities in elected posts, and evidence of discrimination against minorities in other aspects of society that might hinder their ability to participate in the electoral process. Id. at 36-37, 106 S.Ct. at 2758-59. None of these factors are essential to a finding of vote dilution, id. at 45, 106 S.Ct. at 2763-64, but the Supreme Court has explicitly warned that a judgment must encompass not just the three preconditions, but "a comprehensive, not limited, canvassing of relevant facts." De Grandy, supra, ___ U.S. at ___, 114 S.Ct. at 2657. The assessment must be a holistic one.

38 Justice Thomas has highlighted some profound difficulties with the Gingles analysis, Holder, ___ U.S. at ___ — ___, 114 S.Ct. at 2615-18, but this Court is bound to apply it. See, e.g., Patterson v. McLean Credit Union, 491 U.S. 164, 172, 109 S.Ct. 2363, 2370, 105 L.Ed.2d 132 (1989).

Shaw v. Reno provides little guidance for deciding whether an existing districting plan is narrowly tailored to the goal of section 2 compliance. See Shaw, ___ U.S. at ___ — ___, 113 S.Ct. at 2831-32. The logical approach is to begin by determining whether historical, societal, and demographic conditions in Georgia at the time of redistricting sufficed under the Gingles preconditions to reasonably necessitate creating the Eleventh Congressional District. If so, the General Assembly had a compelling interest in designing the plan as it did.

Is the Eleventh Congressional District Required by Section Two?

Any determination of whether three districts were necessary to avoid vote dilution sports prodigious amounts of statistical baggage. The court heard from seven experts on the virtues and vices of the district,³⁹ on topics including compactness, communities of interest, racial polarization in voting, the minimum demographic requirements of a reasonable majority-black district, and Georgia's racist past.⁴⁰

39 We include Ms. Meggers in our count; being a veteran practitioner of redistricting, she is more intimate with the subject than any academic commentator.

40 Traditional districting principles such as compactness, contiguity, communities of interest and respect for political subdivisions are not constitutional requirements, Shaw v. Reno, ___ U.S. at ___, 113 S.Ct. at 2827, but are relevant to our consideration of narrow tailoring. We agree with the North Carolina court that the Supreme Court will probably not

adopt a definition of 'narrow tailoring' in the redistricting context that requires consideration of whether the challenged plan deviates from 'traditional' notions of compactness, contiguity, and respect for political subdivisions to a greater degree than is necessary to accomplish the state's compelling purpose.

Shaw v. Hunt, supra, 861 F. Supp. at 451. Such a standard would elevate to constitutional status that which was intended only as a barometer for determining whether a district adequately serves its constituents. Observance of those traditional principles is also difficult to judge at the exacting level required for a narrow tailoring determination, and such judging would force the judiciary to meddle with legislative prerogatives to an undesirable degree. See id., 861 F. Supp. at 451-53.
Nothing, however, precludes the Court from considering traditional districting principles as guideposts in a narrow tailoring analysis; while not required, they are potentially useful indicators of where the legislature could have done less violence to the electoral landscape. Compactness, for example, is a principle directly relevant to section 2 compliance, which in turn is a central component of our narrow tailoring inquiry.

The more arcane machinations of the parties' statistical experts present us with an initial obstacle. Statistical evidence is potentially of great importance in voting rights cases, e.g., Gingles, 478 U.S. at 52-61, 106 S.Ct. at 2767-72, but the Court has been presented with no particularly dispositive analyses in this case. Drs. Lichtman and Weber — veterans of the Shaw litigation circuit, see Vera v. Richards, No. H-94-0277, 861 F. Supp. 1304 (S.D.Texas 1994), Hays I, 839 F. Supp. at 1208 — launched conflicting assumptions, databases, and bases of measurement, only to be trounced by Dr. Katz, who impugned the testimony of both. The Court, no experts in statistical methodologies and reluctant to base constitutional holdings on numerical niceties,⁴¹ is left with little of use. We deal with each relevant topic as needed, supplementing the expert testimony with case law where the former provides inadequate guidance.

41 Dr. Katz's report reveals some of the potential difficulties in relying on statistical data. See Rpt. at 10.

On Compactness

Our first consideration is geographical compactness — one of the preconditions to a section 2 claim. Unfortunately, there is no litmus test for compactness; it has been described as "such a hazy and ill-defined concept that it seems impossible to apply it in any rigorous sense in matters of law." Tr. IV, at 282. See also Karcher v. Daggett, 462 U.S. 725, 756, 103 S.Ct. 2653, 2673, 77 L.Ed.2d 133 (1983) (stating that compactness requirements have been of limited use because of vague definitions and imprecise application); B. Grofman, Criteria for Districting: A Social Science Perspective, 33 UCLA L.Rev. 85 (1985) (reviewing measures of compactness and stating that none are accepted as definitive). Various mathematically based measures have been created, none of which yielded clear answers in this litigation. See Tr. IV, at 275-288 (testimony of Dr. R. Weber). A few include measuring the ratio of the district's area to the square of the district's perimeter; measuring the ratio of the district's population to the population of the area that would fall inside a rubber band stretched around the district; and measuring the ratio of the district's area to the area of the minimum circle that could circumscribe the district.⁴²

42 Most of these tests are culled from legal and political science scholarship. See, e.g., R. Pildes R. Neimi, Expressive Harms, "Bizarre Districts," and Voting Rights: Evaluating Election District Appearances After Shaw v. Reno, 92 Mich.L.Rev. 483 (1993); E. Roeck, Jr., Measuring Compactness as a Requirement of Legislative Apportionment, 5 Midwest J.Pol.Sci. 70 (1961). See also Karcher, 462 U.S. at 756-58, 756 n. 19, 103 S.Ct. at 2673-74, 2673 n. 19 (Stevens, J., concurring); T. Goldstein, Unpacking and Applying Shaw v. Reno, 43 Am.U.L.Rev. 1135 (1994).

At trial, Dr. Lisa Handley, an expert for the State, presented the innovative if abominably named "meanderingness test" for compactness. This test, formulated by Dr. Handley in conjunction with State's counsel specifically for this litigation, constituted the State's only notable submission on the subject of compactness. The "meanderingness" score measures the irregularity of a district's shape by determining how much of the district can be reached by straight lines emanating from that one point in the district yielding "the highest percentage of direct, straight line accessibility." Rpt. of Dr. Handley, at 8-9. We do not find this test any more useful than the others we have encountered; it measures only a narrow facet of the elusive notion that is "compactness," and it allows for some absurd results. As Plaintiffs noted a trial, this measure would find a pencil-thin district hundreds of miles long 100% compact, since there are no twists or turns to break the straight line measurements. We would add that the "meanderingness test" is especially useless in analyzing the Eleventh District: while the vast — and sparsely populated — core of the Eleventh accounts for the district's favorable score on Dr. Handley's test, the narrow — and densely populated — appendages escape notice. In fact, this test is an excellent means of highlighting the egregiously manipulated portions of any voting district, since those portions will always lie beyond the reach of straight line measurements.

Our criticism of the "meanderingness test" leads us to another genre of compactness considerations: those dealing with population. These measures are far less abstract than those above, and better reveal the deficiencies of the Eleventh Congressional District. They require an assessment of population densities, shared histories, and common interests; essentially whether the populations roped into a particular district are close enough geographically, economically, and culturally to justify them being held in a single district.⁴³

43 This approach is similar to that of "functional" tests for compactness, like that used by Kathleen Wilde in designing the max-black plan. Dillard v. Baldwin County, 686 F. Supp. 1459, 1465 (M.D.Ala. 1988) gives a good example of the functional view:

By compactness, Thornburg does not mean that a proposed district must meet, or attempt to achieve, some aesthetic absolute, such as symmetry or attractiveness. . . . [G]eographical symmetry or attractiveness is . . . a desirable consideration for districting, but only to the extent it facilitates the political process. . . .

[A] district is sufficiently compact if it allows for effective representation. . . . [A] district would not be sufficiently compact if it was . . . so convoluted that there was no sense of community, that is, if its members and its representative could not easily tell who actually lived in the district.

This view of compactness will occasionally conflict with Shaw v. Reno, which places renewed emphasis on actual district shape; post- Shaw, the functional approach may improperly devalue geographical measurements.
Regardless, functionally speaking, there is considerable potential for voter confusion in the current Eleventh District, particularly near the borders in urban areas of the Eleventh. Beyond the general assumption in cities that "if you are black you are in the Eleventh, and if you are white you are not," the erratic lines and split counties and precincts do not afford voters ready indications of the district in which they reside.

A review of the Eleventh District from a population-based perspective confirms that the district is not compact for purposes of section 2 of the VRA. The populations of the Eleventh are centered around four discrete, widely spaced urban centers⁴⁴ that have absolutely nothing to do with each other, and stretch the district hundreds of miles across rural counties and narrow swamp corridors. Two thirds of the population of the district is concentrated in urban DeKalb, Richmond and Chatham counties. These communities are so far apart that DOJ's insistence that they are "compact" renders the term meaningless. The hooks, tails and protrusions of those counties reveal the true "shape" of the district: if it were graphically depicted and sized according to the density of population, the miniature polyp of South DeKalb County would become a large bulbous affair accounting for about 35% of the district's size; the narrow hook into Richmond County would be a rather uncouth polygon sporting about 16% of the district's girth; and the tuft attached to the tail extending to Chatham County would represent 12.4% of the district. Finally, the proboscis extending into Baldwin County would be another distant repository of Eleventh District population.

44 Atlanta, Augusta, Savannah, and Columbus.

As for finding any connection between these discrete clumps of people, Dr. O'Rourke, an expert for the Plaintiffs, presented the Court with compelling evidence of economic conditions, educational backgrounds, media concentrations, commuting habits, and other aspects of life in central and southeast Georgia making it exceedingly clear that there are no tangible "communities of interest" spanning the hundreds of miles of the Eleventh District. Rpt. of Dr. O'Rourke; Tr. III, at 99. Nor would your average citizen think there would be. Assertions to the contrary are the result of shallow and offensive thinking: blacks live in Savannah, blacks live in Atlanta, and so there must be some deep cultural bond between them since, after all, aren't they black?

Far more pernicious has been the Court's willingness to accept the one underlying premise that must inform every minority vote dilution claim: the assumption that the group asserting dilution is not merely a racial or ethnic group, but a group having distinct political interests as well.

Holder, ___ U.S. at ___, 114 S.Ct. at 2597. As far as Gingles' first precondition serves to minimize the effect of this assumption by requiring groups to reside in a compact area, it has done its work today. Blacks in Georgia did indeed rise from a common heritage of slavery and oppression, but with the shedding of those burdens they have begun to follow myriad, distinctive paths. They take their voting preferences with them.

The Eleventh Congressional District, as constructed, does not satisfy the first Gingles precondition, and was not reasonably required by section 2 of the VRA. There simply is no sufficiently large, compact minority population in south-central Georgia warranting creation in that area of a single-member, majority-black district as a vote dilution remedy. Cf. Shaw v. Reno, ___ U.S. at ___, 113 S.Ct. at 2827. For the sake of thoroughness, we now turn to other factors relating to the second and third Gingles pre-conditions.

On Racial Polarization and Overly Safe Districts

We now move from reviewing the need for an Eleventh District to examining the need for the current racial configuration of its constituency. The Court was here subjected to especially mind-numbing and contradictory statistical data. A team of court-appointed statisticians and political scientists would be hard pressed to divine from the evidence a single concrete finding regarding racially polarized voting, crossover voting patterns, or the minimum black voting age population reasonably required under the VRA to ensure "equal opportunity" for black voters in Georgia. Moreover, we are dubious of arguments based on elusive "benchmarks" for the "appropriate" amount of minority voting influence, and of the judiciary's competence to resolve such disputes. The benchmark concept is rife with difficulties, and tolls a dangerous evolution in voting rights jurisprudence. See, e.g., Holder, ___ U.S. at ___, 114 S.Ct. at 2594 ("[I]n setting the benchmark of what 'undiluted' or fully 'effective' voting strength should be, a court must necessarily make some judgments based purely on an assessment of principles of political theory."). We feel, however, constrained to the task.

The "bivariate ecological regression methodologies," "extreme case analyses," "r-squared coefficients," and "frequency distributions" cough up basically two types of statistical information. First, we have analyses of the degree of racial polarization in Georgia voting. Regression and extreme case analyses are the primary research methods here, and Dr. Lichtman, for the United States, uses them to prove the existence of polarization in Georgia. Drs. Weber and Katz also implicitly or explicitly acknowledge the presence of racially polarized voting, but dispute the extent and significance of its existence. All that is clear from the evidence is that some degree of vote polarization exists, but not in alarming quantities. Exact levels are unknowable. Unfortunately, such a finding is not particularly useful, and tells us nothing about the racial percentages needed within a single-member district to "rectify" the problem.

In addition, we also find a significant degree of crossover voting in Georgia and the Eleventh District, with white voters slightly more willing to vote for black candidates than black voters for white candidates. Rpt. of Dr. Katz, Table 2. The average percentage of whites voting for black candidates ranged from 22% to 38%; the average percentage of blacks voting for white candidates ranged from 20%-23%. Black and black-preferred candidates in Georgia have achieved many electoral victories in local and statewide elections and have received significant — occasionally overwhelming — support from both black and white voters within the Eleventh Congressional District. As with our vote polarization findings above, our observations here are not overly useful on their own. However, the very lack of solid evidence of black vote cohesion or rampant bloc voting, i.e., vote polarization, contributes to our conclusion that the current Eleventh District was not an appropriate section 2 remedy.

The second type of information we cull from the available evidence is geared towards determining the actual percentages needed to give black voters in a certain district an equal opportunity to elect their candidate of choice. At first glance it appears that the 60.36% black voting age population of the Eleventh is excessive; the Second and Fifth Districts have 52.33% and 57.47% black voting age population, respectively. Both the Second and Fifth were sufficiently majority-black districts to satisfy DOJ, and so presumably the VRA.

Statistically speaking, reconstituted election results from precincts within a certain district, actual prior election results from a certain district, and frequency distributions are the primary methods used to estimate the percentages needed to give black voters an equal opportunity to elect a candidate of their choice. Exact numbers are not available in this context, and the State did not make any inquiries during its redistricting efforts that might have informed its choice of what black voting population percentage was needed. As Ms. Meggers testified, the "traditional" 65% rule governed: as a rule of thumb, it was thought, 65% black total population was needed in a majority-black district, because approximately 5% + of blacks were not of voting age population, 5% + were unregistered even if of voting age, and some blacks could be expected to not vote, regardless of status. Thus, the black voting strength was "really" slightly over 50%. We will not base a determination of narrow tailoring — a finding of constitutional magnitude — on such a broad generalization.

That said, the expert testimony is extremely sparse on the numbers needed in the Eleventh to ensure equal opportunities for minorities. First the experts examine actual results from the 1992 election in the Eleventh Congressional District. Black candidates were supported in several stages of the election to varying degrees, but the results of a single election provide little basis for sound generalizations. We note that in the 1992 Democratic primary, black candidates in the Eleventh District received approximately 55% of the white vote, signifying to us a general willingness of white voters to vote for black candidates. Cynthia McKinney (who is black), receiving 56% of the vote, won in the runoff election against George DeLoach (who is white). Rep. McKinney received 23% of the white vote in the runoff.

Dr. Lichtman presented reconstituted election analyses involving two elections: the 1988 Democratic presidential primary, and the 1990 Democratic gubernatorial runoff. See Rpt. of Dr. Lichtman at 36. When the results in these two elections are "recompiled" for the present Eleventh Congressional District, black candidate Andrew Young would have prevailed in the gubernatorial runoff by a margin of 58% to 42%, and black candidate Jesse Jackson would have prevailed in the presidential primary by a margin of 68.1% to 31.9%. See id. at 37; Rpt. of Dr. Weber Supp., Attach. F. Dr. Weber's extended reconstituted election analyses indicate that the black candidate(s) of choice have received a total of 65.7% of the vote in ten elections held entirely within the precincts of the current Eleventh Congressional District.

Dr. Joseph Katz presented the final evidence on the percentages necessary to grant black voters equal electoral opportunities. This evidence is actually the most useful to us; the State seems to have retained Dr. Katz for the unusual purpose of undermining the testimony of both Intervenor United States' expert Dr. Lichtman and Plaintiffs' expert, Dr. Weber. He was largely successful. Dr. Katz questioned the assumptions made by Drs. Weber and Lichtman, and his efforts were intended to show that while the current plan did give minorities a "reasonable" opportunity to elect their candidates of choice, the plan was not compelled by the VRA. See Tr. V, at 73; State Def. Prop. Concl. at ¶¶ 108-13. Dr. Katz offered several frequency distributions plotting prior election results to determine the probability of a black candidate winning a future election given a certain percentage of registered black voters. Though the analysis is sound and relatively straightforward, its value is diminished by the use of databases from state-wide elections rather than from specific areas within the state; it cannot account for varying voting patterns in the specific regions pertinent to this case. The analysis also measures only the probability of a black candidate winning, not of a black-preferred candidate winning. Since only black candidates' success is plotted, using the distribution as a predictor requires the unsubstantiated assumption that a black candidate will always be the candidate of choice of black voters in future elections, and ignores the historical preferences of black voters in all-white elections.

Finally, the value of the analysis is lessened by the use of registered black population rather than voting age black population as its independent variable. Voting age population — those persons eligible to vote — is the proper measure of whether black voters have an equal opportunity to elect candidates of their choice, not registered population, which is only that segment of the population that actually decides to participate. See Growe v. Emison, ___ U.S. ___, ___ n. 4, 113 S.Ct. 1075, 1083 n. 4; Marylanders for Fair Representation v. Schaefer, 849 F. Supp. 1022, 1051 (D.Md. 1994) (citing cases). The VRA guarantees "the right to have free and equal access to the ballot box and to have the vote that is cast count the same as any other person's," but the Act "does not endow the voter with the right to have his or her vote cast for the winner." Smith v. Brunswick County, 984 F.2d 1393, 1398 (4th Cir. 1993). Accounting for lower voter registration and turnout rates among black citizens when determining what constitutes an "equal opportunity to participate in the electoral process," Gingles, 478 U.S. at 44, 106 S.Ct. at 2763, and creating "safe black districts" to compensate for those rates amounts to an incentive for and institutionalization of black voter apathy. That we will not condone. As the Supreme Court has recently said, "minority voters are not immune from the obligation to pull, haul, and trade," De Grandy, ___ U.S. at ___, 114 S.Ct. at 2661, and, this Court adds, "the obligation to register and vote." We are troubled, therefore, by analyses founded on black voter registration rather than black voting age population.

Even considering these shortcomings, Dr. Katz's frequency distributions are the most useful and credible guide to estimating the black population percentages needed in the Eleventh Congressional District. Defendants and Intervenors seem to agree that the frequency distributions are the only real basis in evidence for estimating the percentages required, and their proposed conclusions of law on the narrow tailoring issue implicitly or explicitly refer to this data. The data indicates that a black candidate in Georgia would have an "equal" chance of being elected in a district containing 45-50% black registered voters. See Rpt. of Dr. Katz, at Tab 5 (all charts). Georgia's Eleventh Congressional District has a black registered voter percentage of 57%, Def. Joint Exh. 21, providing a black candidate a roughly 73% probability of winning an election in that district. See id. An assessment of black voting age population would presumably yield an even higher probability. Defendants' tacitly acknowledge that this is more than an "equal" chance of election, but they propose that it nonetheless presents a "reasonable" opportunity for blacks to elect a candidate of their choice. We are only concerned with the former assessment: the probability afforded by the current Eleventh District represents more than an "equal" opportunity, and is in excess of that required by section 2 of the VRA.

The record fails to demonstrate compactness, chronic bloc voting, or reasonably necessary black voter percentages in the Eleventh Congressional District. Since the district does not satisfy the Gingles preconditions, we do not review the "totality of the circumstances." The Eleventh District is not narrowly tailored to comply with the putatively compelling interest of section 2 compliance.

In sum, the current districting plan is not reasonably necessary to comply with sections 2 or 5 of the VRA. Since no compelling state interest other than VRA compliance is evident, the plan fails strict scrutiny under the Fourteenth Amendment. We finally conclude and declare that Georgia's Eleventh Congressional District is unconstitutional in its current composition. The State of Georgia is hereby barred from using it in future congressional elections.

III. RELIEF

Elected governments are obviously entrusted with enormous responsibilities, and when confronted with fundamental issues of race and democracy, they are obliged to govern with insight and care. Our prolonged meditations on compactness scores and frequency distributions threaten to obscure the important point that "narrow tailoring to accommodate a compelling state interest" does not mean a wooden, mechanical application of one remedy while ignoring all alternatives. Single-member majority-black districts are not a constitutional or statutory requirement. The assumption that the sole means of enhancing blacks' political influence is to pack them into such districts is unimaginative.

The time has come to contemplate more innovative means of ensuring minority representation in democratic institutions. Otherwise, the United States face a steady transmogrification into racially balkanized voting units. The Court, however, will not lead the pursuit for answers, for we are in agreement with Justice Thomas that

under our constitutional system, this Court is not a centralized politburo appointed for life to dictate to the provinces the "correct" theories of democratic representation, the "best" electoral systems for securing truly "representative" government, the "fairest" proportions of minority political influence, or . . . the "proper" sizes for local governing bodies.

Holder, ___ U.S. at ___, 114 S.Ct. at 2602. Voting reform is the province of legislatures; Georgia's owes its citizens new solutions.

No congressional elections will be held in the Eleventh Congressional District until it is revised in keeping with our decision today. This is a permanent injunction. The Court will initially reconfigure the district with the aid and assistance of the personnel and equipment of the State Reapportionment Services office. Such revision will surely affect the First and Tenth Congressional Districts, but there is little chance that the Court's order will affect other districts in Georgia. The original parties and the Abrams Intervenors are invited to submit written, statistical, and graphic suggestions (not to exceed a total of 25 pages per party) to this Court, no later than September 26, 1994. The parties' submissions should be as narrowly conceived as possible to satisfy constitutional requirements, and cause minimal disruption to the political process of the State of Georgia. A hearing on redistricting remedies will be held in Savannah at 9:00 a.m. on October 3, 1994. After this Court's order for relief as to the 1994 congressional elections, Georgia officials may wish to revisit the issue of reapportionment in the affected districts, or statewide. That decision rests with the State.

CONCLUSION

Accordingly, it is HEREBY ORDERED that congressional elections in Georgia's Eleventh Congressional District are ENJOINED until further order of this Court. It is further ORDERED that this Court reserves decision and jurisdiction to reconfigure the Eleventh Congressional District in a manner consistent with this opinion and after reviewing the parties' suggestions; this Court also reserves decision and jurisdiction to order the concomitant required modifications to the First and Tenth Districts as the need arises.

As an accommodation to the parties and their attorneys, a copy of this opinion shall be lodged with the Clerk of the Eleventh Circuit Court of Appeals simultaneously with its filing in the District.

EDMONDSON, Circuit Judge, dissenting:

At the outset, I cheerfully admit that my colleagues may well be right about what is the correct result in this case. In such circumstances, I (despite my disagreement) usually write nothing in dissent. But, given the exceptional importance of this case to the people of Georgia and to the law of the United States, I feel obliged to explain briefly and generally my view.

Reapportionment of a state's congressional districts is fundamentally the domain of the states rather than the federal courts. Growe v. Emison, ___ U.S. ___, ___, 113 S.Ct. 1075, 1080, 122 L.Ed.2d 388 (1993); Connor v. Finch, 431 U.S. 407, 413-14, 97 S.Ct. 1828, 1833, 52 L.Ed.2d 465 (1977); Reynolds v. Sims, 377 U.S. 533, 585-87, 84 S.Ct. 1362, 1394, 12 L.Ed.2d 506 (1964). The Constitution leaves the responsibility for apportioning congressional districts with the states. Id; Voinovich v. Quiltner, ___ U.S. ___, ___, 113 S.Ct. 1149, 1157, 122 L.Ed.2d 500 (1993). And, state political considerations pervade the apportionment process making it a "highly complex and multifaceted . . . political thicket." Reynolds, 377 U.S. at 568-69, 84 S.Ct. at 1385. For example, considerations of incumbency, local politics, and the balance of power between Democrats and Republicans in the state legislature, as well as innumerable other considerations (such as, the simple exchange of votes by legislators to advance each other's goals), control the placement of a state's congressional districts. All legislative action is a product of debate, compromise, and political decisions. The evidence — such as the Lieutenant Governor's testimony — in this case shows that apportionment legislation takes this process to an especially intense level.

Largely because of the uniquely political, sensitive, and complex nature of apportionment, a federal court hearing a challenge to a state's apportionment plan must give federalism concerns especially significant consideration. The Supreme Court has repeatedly stressed that "[f]ederal courts are barred from intervening in state apportionment in the absence of a violation of federal law precisely because it is the domain of the states, and not the federal courts, to conduct apportionment in the first place." Voinovich, ___ U.S. at ___, 113 S.Ct. at 1157. Federal courts are bound to respect a state's apportionment decisions unless the choices have clearly violated constitutional or federal law requirements. Id. ¹

1 Plaintiffs in this case attempt to minimize federalism concerns by suggesting that the federal government, acting through the Department of Justice, forced the State of Georgia, that is, the State's elected leaders, to enact the present congressional districting plan. Because Georgia is subject to the preclearance requirements of section 5 of the Voting Rights Act, the Department of Justice does play a role in the enactment of any State plan. The State, however, always has the option of seeking relief in federal court from the Department's objections. State leaders bear the responsibility for weighing the advantages and disadvantages of litigation as part of their responsibility for reapportionment generally. But, that the courts are open to Georgia undercuts substantially the idea that the Department of Justice can, in fact, dictate to Georgia. (I note, by the way, that Georgia's Senate had included parts of Savannah — the most controversial element of the Eleventh District — in the Eleventh District before the Department of Justice mentioned such a thing.) In addition, no plan may be enacted unless it is first adopted by Georgia's legislature. The Department's objection to the State's first two plans affected Georgia's ultimate plan. But in the end, the plan before us represents the judgment of Georgia's elected leaders (the main guardians of the public interest for Georgia) on congressional apportionment. And, as the Supreme Court has stated, the State's judgment demands the respect of the federal courts.

With these federalism concerns in mind, I look at Shaw v. Reno, ___ U.S. ___, 113 S.Ct. 2816, 125 L.Ed.2d 511 (1993), the precedent that allows, but also limits in important ways, federal judicial action in apportionment cases. Because Congressional apportionment has always been seen as a state matter, only two circumstances historically would allow a federal court to use the Equal Protection Clause as a basis to intervene in the apportionment process: (1) the state's apportionment plan violated one-person, one-vote principles by creating districts with substantially unequal populations, Reynolds v. Sims, 377 U.S. 533, 84 S.Ct. 1362, 12 L.Ed.2d 506 (1964); (2) the state's apportionment plan had the purpose and effect of diluting the voting strength of a minority or other identified group of voters, Wright v. Rockefeller, 376 U.S. 52, 84 S.Ct. 603, 11 L.Ed.2d 512 (1964). Neither of these circumstances is present in this case.

Therefore, if we are to interfere with Georgia's Congressional plan, from where does our authority come? Shaw points to the possible source. Shaw held that a congressional district that is "so highly irregular on its face that it can be viewed only as an effort to segregate the races for purposes of voting" may be examined by federal courts and, then, be struck down unless the district can survive "strict scrutiny" under the Equal Protection Clause. Shaw, ___ U.S. at ___, 113 S.Ct. at 2828. Highly irregular shape — the appearance of the district — is the critical element of this new cause of action under the Equal Protection Clause. This element strongly limits the power of federal courts to become entangled in state apportionment matters.

Plaintiffs in this case want us not just to follow Shaw, but to extend Shaw to make federal court intrusion in apportionment cases easier and more likely. But, there are important differences between following a precedent and extending it. Because extension of Shaw as plaintiffs advocate would diminish in important ways the traditional freedom of states to apportion themselves for Congressional purposes without federal judicial supervision, I decline to take that step without plain instruction from the nation's highest court.

Plaintiffs argue that the fundamental inquiry is whether the district has been set up on racial grounds. The burden, they admit, is on them to show that legislators intentionally drew lines to segregate voters according to race. Plaintiffs claim, however, that any evidence probative of the state's decision-making process may be used to prove legislative intent. Under this more active approach, the district's appearance is of little importance; shape is merely one piece of circumstantial evidence proving the legislature's intentional use of race. Other evidence, especially direct testimony from legislators about their motivations, could be used by plaintiffs to prove intent.² According to plaintiffs, once they, by any means, have shown that race was a substantial factor in the legislature's decision-making process, the burden is on the state to show that the district satisfies strict scrutiny. I cannot agree with this more sweeping approach to federal judicial review. The words "on its face" in Shaw were not used lightly, I believe.³ The district's shape is no mere piece of evidence; it is a critical part of the cause of action.

2 During the trial, several State legislators testified about why the Eleventh looks like it does. I accept that each of these men spoke the truth as they know it. But, the main truth (with which I think all agreed) is that, in a democracy, apportionment is a highly political process. Many different agendas may be at work. So, no one person can truly say why other legislators decided to vote this way or that way. I cannot give much weight to post-enactment statements of individual legislators about the subjective motives of the State house, State senate and governor when the apportionment plan was enacted. See generally Edwards v. Aguillard, 482 U.S. 578, 595-96 n. 19, 107 S.Ct. 2573, 2584 n. 19, 96 L.Ed.2d 510 (1987) (finding post-enactment statements "to be of little relevance in determining the intent of the legislature"); Blanchette v. Connecticut General Insurance Corp., 419 U.S. 102, 132-33, 95 S.Ct. 335, 353, 42 L.Ed.2d 320 (1974) (holding that statements by individual legislators "represent only the personal views of those legislators" and not the intent of the legislature); National Woodwork Mfrs. Ass'n. v. N.L.R.B., 386 U.S. 612, 639 n. 34, 87 S.Ct. 1250, 1265 n. 34, 18 L.Ed.2d 357 (1967) (holding that statements made by legislators after a bill was passed "could represent only the personal views of those legislators").
Shaw's highly-irregular-appearance standard is an objective standard, and it, among other things, avoids the unreliability of after-the-fact statements by legislators and also avoids the necessity of federal judges probing and intruding on state officers' thoughts (perhaps expressly crediting some state officers and not crediting others — a divisive event) to "learn" the state's true intent.

3 No fewer than six times the Shaw Court repeats that a district may be challenged on the grounds that it is "bizarre," "highly irregular" or "irrational on its face." Id. at ___, ___, ___, ___, ___, 113 S.Ct. at 2824, 2825, 2826, 2829, 2832. That the Court presented the issue in this manner cannot be ignored. To hold that the proper inquiry is simply whether race was a "substantial factor" ignores the plain language of the opinion. Instead, to follow Shaw, a court must ask whether the district is so bizarre on its face that only racial considerations could account for its shape.
And, I consider the Court's discussion of United Jewish Organizations of Williamsburg, Inc. v. Carey, 430 U.S. 144, 97 S.Ct. 996, 51 L.Ed.2d 229 (1977) (UJO), enlightening. In UJO, the Court upheld a state's districting scheme despite evidence that the lines were drawn on the basis of race. According to Shaw, the plaintiffs in UJO failed to state a valid claim because they "did not allege that the plan, on its face, was so highly irregular that it could be understood only as an effort to segregate voters by race." Shaw, ___ U.S. at ___, 113 S.Ct. at 2829. Again, appearances count a lot: Is the district highly irregular in its shape?

In turning to the facts of this case, we must ask first whether the plaintiffs have carried their burden of showing that the Eleventh District is so "bizarre" or "highly irregular" in shape that it can only be explained by race. In determining whether a district is bizarre, the court may consider many factors, including the district's size, outline, allegiance to traditional districting principles such as compactness and respect for political sub-divisions, and how it appears in comparison to other districts in the State's plan, earlier plans, or even districts from other states. Based on objective criteria, I cannot find and cannot conclude that plaintiffs have proved the Eleventh District is bizarre or highly irregular within the meaning of Shaw.

1. The size of the district is not particularly noteworthy. Georgia's Districts One, Two, and Eight each have a total area of over 10,100 square miles. In contrast, the Eleventh District has an area of 6,780 square miles.

2. The district's 1,184 miles of borders is not distinctive when compared to the Second (1243 miles) or the Eighth (1155 miles) Districts.

3. Plaintiffs claim that the district is bizarre because it comes within 58 miles of crossing the entire State. But, the Ninth District spans the entire northern border of the State and the First, Second, and Eighth Districts begin at the Florida border and stretch north to almost the middle of the State.

4. The Eleventh District shows considerable respect for existing political boundaries. That seventy-one percent of the district's boundaries follow existing state, county, and city borders is significant. This places the Eleventh at the average for the State's ten other congressional districts.

5. In addition, eighty-three percent of the Eleventh's area comes from whole counties. In comparison, the average among the State's other districts is sixty-two and one-half percent.

6. Georgia's congressional districts have no tradition of being neat, geometric shapes. And, areas of the Second and Eighth Districts in Bibb and Houston Counties look — as irregular or — much more irregular. The same may be said of the Third District in Crawford County. To be sure, the Eleventh makes curious turns in some areas, particularly the areas around Atlanta, Savannah, and Augusta. But, in these areas most of the lines follow existing city boundaries or major highways and roads. The Eleventh District makes use of what some have called "landbridges" to connect populations to the district, but it is not unique in that respect: for example, see the current Seventh District's inclusion of Marietta and the 1970's Seventh District's inclusion of an area in or around Dalton.⁴ A comparison of — including, just looking at — Georgia's Eleventh District to the district in Shaw or to districts being successfully challenged in other states supports the conclusion that the Eleventh District is not highly irregular or, put differently, bizarre.⁵

4 About population, I do not say that population distribution within a district counts for nothing when courts look at the district's appearance for Shaw purposes, but I am unsure that the "on its face" rule includes population within the district. In any event, population density normally varies considerably across most districts, except completely urban districts. I expect that many districts (including, for example, Georgia's Ninth District which is, by the way, mostly white) would look funny if geography were somehow distorted — parts blown-up or shrunk — to reflect population density.

5 In thinking about the degree of irregularity in the shape of the current Eleventh District I have also looked at some proposals for apportionment that were never adopted. Many groups submitted proposals for reapportionment on the basis of one person, one vote. One plan was (and is) known as the "Max-Black" plan or the "Max Plan". The proposal was drawn with one idea in mind: to maximize black representation in Congress and as a corollary to boost black voting power in certain districts in Georgia. The plan shows what the Eleventh District would look like if it was wholly based on race. The Max-Black plan did influence to some degree the shape of the ultimate Eleventh District, a majority black district. But, it seems important to me that the actual Eleventh is not identical to the Max-Black plan. The Eleventh, to my eye, is significantly different in shape in many ways. These differences show, as I understand it, consideration of other matters beyond race, including traditional districting factors (such as keeping political subdivisions intact) and the usual political process of compromise and trades for a variety of nonracial reasons. In the light of Shaw, the Max-Black proposal for what became the Eleventh District would likely have been bizarre in shape, but the actual Eleventh District is much more regular. See Appendix A.

7. That the Eleventh District splits eight counties is unremarkable. The Sixth District has no whole counties but parts of five counties. And, the Fourth District has one whole county and parts of two others. Splitting counties in Georgia's congressional districts for reasons wholly unrelated to race is part of Georgia's history: for example, Whitfield in 1970 and Gwinnett in 1980.

8. Qualitative measurements for compactness, such as perimeter and dispersion measurements, show the Eleventh District is not bizarre or highly irregular. One measurement, called dispersion scoring, compares the area of the district to the area of the smallest circle that would enclose the district. And, perimeter scoring compares the area of the district to the area of a circle with the same perimeter. While I express no view as to the reliability of these measurements, political scientists, according to plaintiffs' experts, use the measurements to evaluate a district's compactness.

According to perimeter scoring, the Eleventh is more compact than approximately forty-six other congressional districts in the country. Under dispersion measurements, the Eleventh is more compact than about twenty-nine other districts. And, under these measurements, the Eleventh is more compact than the North Carolina District challenged in Shaw, the Texas and Louisiana Districts recently held unconstitutional in the Fifth Circuit, and the Florida District under challenge in the Eleventh Circuit. Again, these measurements, by themselves, are not outcome determinative; nonetheless, they confirm the view that Georgia's Eleventh District is not highly irregular.⁶

6 One of plaintiffs' expert witnesses testified that under these measurements, as well as population compactness measurements, the Eleventh District scored above traditional cutoffs for compactness.

In concluding, I will not be coy: no party in this case disputes that race was an important consideration in the drawing of the Eleventh District. But, as I understand the law as explained by the Supreme Court, the Constitution does not condemn all race-conscious districting. For Shaw, appearance is the first and the main issue. Highly irregular appearance must be shown, or federal judges have no right to look further or to inquire more.

Federal courts must be on guard against the temptation to venture into the realm of political science. And not every district that some expert or some federal judge could draw "better," that is, more regularly or more compactly, is "highly irregular" within the meaning of Shaw. Whatever I may personally think about the wisdom of the Eleventh's boundaries or of the policy choices that led to them, the Eleventh District is not so bizarre on its face to justify, pursuant to the Constitution, a more searching inquiry. Thus, plaintiffs have failed to prove a valid Shaw claim. In the absence of such a showing, this court — as I understand the law — cannot rightly interfere with the reapportionment plan that resulted from Georgia's political process.

I would render a judgment for defendants.