Larios v. Cox ( 2004 )

314 F.Supp.2d 1357 (2004)

Sara LARIOS, et al., Plaintiffs,
v.
Cathy COX, Defendant.

No. CIV.A.1:03 CV 693-CAP.

United States District Court, N.D. Georgia, Atlanta Division.

April 15, 2004.

Frank B. Strickland, Anne Ware Lewis, Strickland Brockington Lewis, Stacy Grant Freeman, Arnall Golden & Gregory, Atlanta, GA, E. M. Braden, Amy M. Henson, Baker & Hostetler, Washington, DC, for Plaintiffs.

Dennis Robert Dunn, Thurbert E. Baker, Office of State Attorney General, Mark Howard Cohen, Troutman Sanders, David F. Walbert, Parks Chesin & Walbert, Atlanta, GA, for Defendants.

Before MARCUS, Circuit Judge, PANNELL and O'KELLEY, District Judges.

PER CURIAM.

On March 25, 2004, this Court issued an Order adopting and implementing a set of interim redistricting plans for the Georgia General Assembly ("Special Master's 1-B Plans"). To ensure that the Secretary of State and the election officials of the State of Georgia could fulfill their duties timely, accurately, and in accordance with the law, the Court entered an abbreviated Order at that time. This more detailed Memorandum Opinion follows and explains that Order.

I.

On February 10, 2004, after trial, this three-judge Court ruled in favor of the plaintiffs on their claim that the existing reapportionment schemes for the Georgia Senate and House of Representatives violated the one person, one vote principle firmly rooted in the Fourteenth Amendment's Equal Protection Clause. We found that the existing 2001 House and 2002 Senate plans ("Enjoined Plans") arbitrarily and discriminatorily diluted and debased the weight of certain citizens' votes by intentionally underpopulating districts in rural south Georgia and by correspondingly overpopulating the districts in the suburban areas surrounding Atlanta. We therefore enjoined the defendant from any further use of those plans in future elections, and we urged the Georgia General Assembly to adopt newly enacted reapportionment plans that would be acceptable to the legislature and the Governor and conform to the Constitution and laws of the United States. Larios v. Cox, 300 F.Supp.2d 1320, 1356-58 (N.D.Ga.2004). The Georgia General Assembly was unable to meet the March 1, 2004 deadline and as yet has submitted no new plans for review by the Court or the Department of Justice. Accordingly, it fell to this Court to draw interim reapportionment plans for use in time for the orderly conduct of the November 2004 elections. See Wise v. Lipscomb, 437 U.S. 535, 540, 98 S.Ct. 2493, 2497, 57 L.Ed.2d 411 (1978); Reynolds v. Sims, 377 U.S. 533, 585-87, 84 S.Ct. 1362, 1393-94, 12 L.Ed.2d 506 (1964).

By Order dated March 1, 2004, we appointed Mr. Joseph Hatchett to serve as Special Master pursuant to Rule 53 of the Federal Rules of Civil Procedure. Thereafter, on March 2, 2004, we adopted a series of guidelines to inform the Special Master in the process of preparing reapportionment maps for the Senate and the House of Representatives of the General Assembly of Georgia ("March 2 Guidelines"). As we noted at the time, preparing reapportionment plans in a timely manner, while reconciling the demands of the Constitution, the Voting Rights Act, and the redistricting principles traditionally recognized by the State of Georgia, presented a substantial undertaking. Indeed, the reapportionment of legislative bodies is "a legislative task which the federal courts should make every effort not to pre-empt." Wise, 437 U.S. at 539, 98 S.Ct. at 2497. However, where a federal court has declared an existing apportionment scheme unconstitutional and the legislature, after being afforded a reasonable opportunity, fails to adopt a substitute plan, "it becomes the `unwelcome obligation' ... of the federal court to devise and impose a reapportionment plan pending later legislative action." Id. at 540, 98 S.Ct. at 2497 (quoting Connor v. Finch, 431 U.S. 407, 415, 97 S.Ct. 1828, 1834, 52 L.Ed.2d 465 (1977)).

The Court's guidelines established three principal criteria for drafting new Senate and House Plans: the Constitution, the Voting Rights Act, and certain traditional and neutral principles of redistricting. Plainly, the requirements of the Constitution and the Voting Rights Act took precedence over any traditional redistricting principles. Arizonans for Fair Representation v. Symington, 828 F.Supp. 684, 687 (D.Ariz.1992) (three-judge court); see also Good v. Austin, 800 F.Supp. 551, 554 (E.D. Mich.1992).

Because the core constitutional wrong to be remedied in this case was a violation of the Fourteenth Amendment's one person, one vote principle, equality of population was a paramount concern in redrawing the maps. See Colleton County Council v. McConnell, 201 F.Supp.2d 618, 627 (D.S.C.2002) (three-judge court). The Equal Protection Clause, as we noted, requires that "the seats in both houses of a bicameral state legislature must be apportioned on a population basis." Reynolds, 377 U.S. at 568, 84 S.Ct. at 1385. "[T]he overriding objective must be substantial equality of population among the various districts, so that the vote of any citizen is approximately equal in weight to that of any other citizen in the State." Id. at 579, 84 S.Ct. at 1390.

The next major priority in drafting the plans was to ensure full compliance with the Voting Rights Act, 42 U.S.C. § 1973. While the preclearance requirement of § 5 does not apply to reapportionment plans prepared and adopted by federal courts to remedy constitutional violations, Abrams v. Johnson, 521 U.S. 74, 95, 117 S.Ct. 1925, 1938, 138 L.Ed.2d 285 (1997); Connor v. Johnson, 402 U.S. 690, 691-92, 91 S.Ct. 1760, 1762, 29 L.Ed.2d 268 (1971) (per curiam), the courts should nonetheless comply with the racial-fairness mandates of § 2 of the Act, as well as the purpose-or-effect standards of § 5 of the Act. See Abrams, 521 U.S. at 90, 96, 117 S.Ct. at 1935, 1938; see also Colleton, 201 F.Supp.2d at 628 (expressing a similar rule); Smith v. Clark, 189 F.Supp.2d 529, 539-40 (S.D.Miss.2002) (three-judge court) (same). Accordingly, the Special Master was directed to ensure that the plans neither diluted voting strength on the basis of race, color, or membership in a language minority group, Thornburg v. Gingles, 478 U.S. 30, 106 S.Ct. 2752, 92 L.Ed.2d 25 (1986), nor led to retrogression in the position of racial minorities, Beer v. United States, 425 U.S. 130, 96 S.Ct. 1357, 47 L.Ed.2d 629 (1976).

The final set of principles governing the drafting process was secondary to ensuring compliance with the Constitution and the Voting Rights Act. We recognized and directed the Special Master to apply Georgia's traditional redistricting principles of compactness, contiguity, minimizing the splits of counties, municipalities, and precincts, and recognizing communities of interest. See Abrams, 521 U.S. at 84, 117 S.Ct. at 1932-33 (affirming Johnson v. Miller, 922 F.Supp. 1556 (S.D.Ga.1995) (three-judge court)); see also Guidelines Adopted by the House Committee on Congressional and Legislative Reapportionment and Redistricting 1991-1992, at 1-2 ("1991 Guidelines"); 2001-02 Guidelines for the House/Senate Committee on Congressional and Legislative Reapportionment and Redistricting, at 2-3 ("2001 Guidelines"). We also directed the Special Master to not take into account the "purely political considerations that might be appropriate for legislative bodies." Wyche v. Madison Parish Police Jury, 635 F.2d 1151, 1160 (5th Cir. Unit A Feb.1981).^[1] We strictly prohibited the Special Master and his experts and assistants from reviewing or analyzing political data and information, including, but not limited to, prior districts' voting performance, incumbent residency, political party registration and past election results. Finally, we observed that "a court-drawn plan should prefer single-member districts over multimember districts, absent persuasive justification to the contrary." Wise, 437 U.S. at 540, 98 S.Ct. at 2497; see also East Carroll Parish Sch. Bd. v. Marshall, 424 U.S. 636, 639, 96 S.Ct. 1083, 1085, 47 L.Ed.2d 296 (1976) (per curiam); Connor v. Johnson, 402 U.S. at 692, 91 S.Ct. at 1762.

On March 15, 2004, the Special Master filed with the Court a Report and Recommendation proposing the adoption of new plans ("Special Master's Original Plans") for the Georgia Senate and House of Representatives.^[2] In developing the redistricting plans for the Senate and the House, the Special Master and his assistants had available to them the entire record of the proceedings in this litigation.

In addition, as the Special Master observed, the state of Georgia and its various political subdivisions have been party to many cases addressing redistricting and related issues, including Georgia v. Ashcroft, 539 U.S. 461, 123 S.Ct. 2498, 156 L.Ed.2d 428 (2003); Abrams, 521 U.S. 74, 117 S.Ct. 1925; Miller v. Johnson, 515 U.S. 900, 115 S.Ct. 2475, 132 L.Ed.2d 762 (1995); Holder v. Hall, 512 U.S. 874, 114 S.Ct. 2581, 129 L.Ed.2d 687 (1994); Johnson v. Hamrick, 296 F.3d 1065 (11th Cir.2002); Clark v. Putnam County, 293 F.3d 1261 (11th Cir.2002); DeJulio v. Georgia, 290 F.3d 1291 (11th Cir.), cert denied, 537 U.S. 948, 123 S.Ct. 413, 154 L.Ed.2d 293 (2002); and Johnson v. Hamrick, 196 F.3d 1216 (11th Cir.1999). These appellate cases were often preceded and succeeded by several district court cases, including Smith v. Cobb County Bd. of Elections and Registrations, 230 F. Supp.2d 1313 (N.D.Ga.2002); Johnson v. Hamrick, 155 F. Supp.2d 1355 (N.D.Ga.2001), aff'd, 296 F.3d 1065 (11th Cir.2002); Johnson v. Hamrick, No. 2:91-CV-002-WCO, 1998 WL 476186 (N.D.Ga. June 10, 1998), rev'd, 196 F.3d 1216 (11th Cir.1999); Johnson v. Miller, 864 F.Supp. 1354 (S.D.Ga.1994) (three-judge court), aff'd, 515 U.S. 900, 115 S.Ct. 2475, 132 L.Ed.2d 762 (1995); and Johnson v. Miller, 922 F.Supp. 1556 (S.D.Ga.1995) (three-judge court), aff'd sub nom. Abrams v. Johnson, 521 U.S. 74, 117 S.Ct. 1925, 138 L.Ed.2d 285. To the extent those opinions contained background information, they provided additional source material. Finally, by Order dated March 2, 2004, the parties were directed to file jointly the record in the Georgia v. Ashcroft proceedings in the District Court of the District of Columbia. [Doc. # 194]. The Special Master, therefore, also reviewed those materials to the extent deemed appropriate.

On March 19, 2004, soon after the Special Master filed with the Court a Report and Recommendation, including proposed redistricting plans for the Georgia General Assembly, interested parties filed comments and objections to the Plans. See Doc. Nos. 197-201 & 204-205 (collectively, "March 19 Comments").^[3]

Thereafter, the Special Master reviewed the objections, including concerns raised about incumbency protection, and, pursuant to subsequent court orders, for the first time, reviewed data concerning incumbents. The Special Master was instructed to do so only pursuant to very stringent and unambiguous guidelines. First, he was permitted to review the issue of incumbency protection only as a distinctly subordinate consideration — subordinate to the Constitution and the mandate of one person, one vote, the Voting Rights Act, and the traditional state interests of compactness, contiguity, minimizing the splits of counties, municipalities, and precincts, recognizing communities of interest, and avoiding multi-member districts. Moreover, he was directed to not consider any change if it would cause any degradation of the original guidelines or the ensuing plans. Finally, the Special Master was instructed to review incumbency protection in a wholly consistent, neutral and impartial manner.^[4]

We permitted the Special Master to consider political incumbency, albeit only in this way, because the Supreme Court has plainly said that this is a legitimate state interest that may be properly considered by a court as well as a legislative body. See, e.g., Abrams, 521 U.S. at 84, 117 S.Ct. at 1933 (affirming district court's redistricting plan which subordinated but considered the avoidance of incumbent contests); Bush v. Vera, 517 U.S. 952, 964, 116 S.Ct. 1941, 1954, 135 L.Ed.2d 248 (1996) (recognizing incumbency protection, at least in the form of avoiding contests between incumbents, as a legitimate state goal); White v. Weiser, 412 U.S. 783, 795-97, 93 S.Ct. 2348, 2354-56, 37 L.Ed.2d 335 (1973) (reversing district court when it chose plan that disregarded traditional state policy of avoiding incumbent contests over alternative plan that accommodated this interest). Moreover, the Supreme Court has held that "a federal district court, in the context of legislative reapportionment, should follow the policies and preferences of the State, as expressed in statutory and constitutional provisions or in the reapportionment plans proposed by the state legislature, whenever adherence to state policy does not detract from the requirements of the Federal Constitution." White, 412 U.S. at 795, 93 S.Ct. at 2355. Georgia's state legislature has historically followed the principle of avoiding contests between existing incumbents. See Johnson v. Miller, 922 F.Supp. at 1565, aff'd sub nom. Abrams, 521 U.S. at 84, 117 S.Ct. at 1932-33; see also 1991 Guidelines.

It was of particular importance that the Special Master consider this subordinated interest in a neutral and consistent way because one of the problems we found in the Enjoined Plans was that population deviations were "systematically and intentionally created ... to protect Democratic incumbents." Larios, 300 F.Supp.2d at 1338. The principle of incumbency protection in the 2002 plans was applied in a "blatantly partisan and discriminatory manner .... [B]oth the House and Senate Plans actually pitted numerous Republican incumbents against one another, while generally protecting their Democratic colleagues." Id. at 1347. Thus, our strict instructions to the Special Master ensured that the unpairing of incumbents would not "validate the very maneuvers that were a major cause of the unconstitutional districting." Abrams, 521 U.S. at 86, 117 S.Ct. at 1933.

Thereafter, on March 22, 2004, the Special Master filed a Supplement to Report and Recommendation of the Special Master ("Supplement"), along with the Alternate Plans, responding to the concerns raised by the March 19 Comments. On March 22, 2004, we conducted an extensive hearing during which comments and objections were lodged by a variety of interested persons and parties, including the plaintiffs, the defendant, Robert S. Kahn and the Democratic Party of Georgia ("Democratic Party"), the Georgia Black Legislative Caucus ("GBLC"), and the Democratic Caucus of the House of Representatives ("Democratic Caucus").^[5] With leave of court, a number of the interested parties were granted additional time to address the Special Master's second alternative set of plans. Those comments were received by the Court and the Special Master on March 23, 2004.^[6] The Special Master, in response to those comments and suggestions, prepared still a third set of plans for the General Assembly ("Special Master's 1-B Plans") and a final Report and Recommendation ("Special Master's 1-B Report").

Finally, on March 25, 2004, we held another hearing^[7] and adopted the Special Master's 1-B Plans as the Court's interim reapportionment plans for the Georgia General Assembly.^[8]

II.

After thorough review of the voluminous record before us, we are satisfied that the Special Master's 1-B Plans comply fully with the Constitution and the principle of one person, one vote, the Voting Rights Act, and the traditional redistricting guidelines of compactness, contiguity, minimizing the splits of counties, municipalities, and precincts, recognizing communities of interest, and avoiding multi-member districts. We are also completely satisfied that the Special Master considered comments and objections concerning political incumbency as only a subordinated and secondary consideration and in full compliance with the strict guidelines we imposed upon him.

To begin, the core violation we found in the Enjoined Plans was the intentional and systematic failure to adhere to the one person, one vote command of the Equal Protection Clause. The Special Master's 1-B Plans substantially resolve that problem by significantly lowering the population deviations found in each voting district.

Based on the 2000 Census, the ideal size for a Senate district is 146,187 persons and the ideal size for a House district is 45,480 persons. The Special Master's 1-B Senate plan ("1-B Senate Plan") has a total population deviation of 1.93% and an average deviation of 0.56%. The 1-B Senate Plan districts deviate from the ideal equal population by a range of +0.95% to -0.99%,^[9] with the largest district having 147,570 persons and the smallest district having 144,745 persons. In sharp contrast, the enjoined 2002 Senate Plan ("Enjoined Senate Plan") contained a total population deviation range of 9.98% and an average deviation of 3.78%. The Enjoined Senate Plan districts deviated from ideal equal population by a range of +4.99% to -4.99%, with the largest district having 153,489 persons and the smallest district having 138,894 persons. Indeed, in the Enjoined Senate Plan, thirty-seven of the fifty-seven districts (66.07%) had population deviations greater than ±4%. Thirty-one districts (55.36%) had deviations greater than ±4.5%, and sixteen districts (28.57%) had deviations greater than ±4.9%.

The Special Master's 1-B House plan ("1-B House Plan") has a total population deviation range of 1.95% and an average deviation of 0.47%. The 1-B House Plan districts deviate from the ideal equal population by a range of +0.97% to -0.99%,^[10] with the largest district having 45,921 persons and the smallest district having 45,032 persons. In stark contrast, the enjoined 2001 House Plan ("Enjoined House Plan") contained a total population deviation range of 9.98% and an average deviation of 3.47%. The Enjoined House Plan districts deviated from ideal equal population by a range of +4.99% to -4.99%, with the largest district having 176,939 persons (in a four-member district) and the smallest district having 43,209 persons. Notably, ninety of the 180 House seats (50%) were in districts with population deviations greater than ±4%. Sixty seats (33.33%) were in districts with deviations greater than ±4.5% and twenty seats (11.11%) were in districts with deviations greater than ±4.9%.

Moreover, and equally notable, the geographic distribution of deviations contained in the Special Master's 1-B Plans contrasts sharply with that of the Enjoined Plans, where the most underpopulated areas were intentionally located almost exclusively in rural south Georgia and inner-city Atlanta, and the most overpopulated areas were located by design almost exclusively in the areas of north Georgia that encircle Atlanta. In the Special Master's 1-B Plans, the ±1% deviations are randomly scattered across Georgia.

Second, the Special Master's 1-B Plans completely fulfill the mandates embodied in §§ 2 and 5 of the Voting Rights Act, 42 U.S.C. §§ 1973 & 1973c. The Special Master's 1-B Plans comply with § 2 of the Voting Rights Act, which provides that "[n]o voting qualification or prerequisite to voting or standard, practice, or procedure shall be imposed or applied by any State or political subdivision in a manner which results in a denial or abridgement of the right of any citizen of the United States to vote on account of race or color." 42 U.S.C. § 1973(a). They also meet and exceed the standards of § 5, which prohibits voting-procedure changes that would result in "a retrogression in the position of racial minorities with respect to their effective exercise of the electoral franchise." Beer, 425 U.S. at 141, 96 S.Ct. at 1364.

A § 2 violation is not established unless, "based on the totality of circumstances, it is shown that the political processes leading to nomination or election in the State or political subdivision are not equally open to participation by members of [a protected class of citizens] 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." 42 U.S.C. § 1973(b). The Special Master's 1-B Plans clearly provide all citizens an equal opportunity to participate in the political process. Measured by voting age population in 2000, African-Americans comprise 26.72% of Georgia's population; measured by registered voters in 2000, African-Americans comprise 25.62% of Georgia's population. In Senate Plan 1-B, 23.20% of the districts are majority-minority; in House Plan 1-B, 24.44% of the districts are majority-minority. Accordingly, "no violation of § 2 can be found here, where ... minority voters form effective voting majorities in a number of districts roughly proportional to the minority voters' respective shares in the voting-age population." Johnson v. De Grandy, 512 U.S. 997, 1000, 114 S.Ct. 2647, 2651, 129 L.Ed.2d 775 (1994).^[11]

As for § 5, the Special Master's 1-B Plans evidence no retrogression from the last valid plans in effect and used for the 2000 elections ("Benchmark Plans").^[12] Indeed, the Special Master's 1-B Plans substantially exceed the number of majority-minority districts created in the Benchmark Plans. The 1-B Senate Plan has thirteen majority-minority districts, the same number contained in the Senate plan used in the 2000 election ("Senate Benchmark Plan").^[13] In addition, the 1-B Senate Plan increases the number of influence districts with 30% to 49% African-American Registered Voters from the eight districts in the Senate Benchmark Plan to the ten districts in the 1-B Senate Plan.^[14] The 1-B House Plan actually increases by about 13% the number of majority-minority districts, as compared to the 2000 House plan ("House Benchmark Plan"). While the House Benchmark Plan had thirty-nine majority-minority districts, the 1-B House Plan has forty-four majority-minority districts.^[15] Thus, the Special Master's 1-B Plans do not cause any retrogression in the position of racial minorities.

Notably, both the amicus GBLC and the Democratic Party agreed that the Special Master's 1-B Plans completely adhere to the Voting Rights Act principles embodied in § 2 and § 5.^[16] With respect to the Voting Rights Act, the defendant objected to the plans principally on these grounds: first, the Special Master supposedly relied primarily on Black Registered Voters ("BREG") over Black Voting Age Population ("BVAP") to measure the retrogressive effect of the plans; and second, the Special Master's 1-B Plans supposedly decreased the number of districts with a BVAP between 25% and 50%.

The defendant first argues that the Special Master improperly relied upon the population of BREG, as opposed to BVAP, in determining compliance with the Voting Rights Act. The defendant relies on several cases from the Eleventh Circuit and the Supreme Court, none of which support her position.

In Solomon v. Liberty County, 899 F.2d 1012 (11th Cir.1990) (en banc), the Court considered a § 2 challenge to the county commission and the school board reapportionment plans in Liberty County, Florida. The county was divided into five districts, and the entire county voted for one representative from each district. Blacks comprised 11% of the total population of the county and were primarily located in District 1. All of the black candidates who had previously run had been unsuccessful. The en banc Court was divided on certain issues in the Solomon case, but both concurring opinions rejected the district court's use of black voting age population instead of black registered voter population in considering the first Gingles factor, i.e. whether blacks constituted a group sufficiently large and geographically compact that they would have the potential to elect their own representative from a single-member district.^[17]

The Solomon Court's preference for voting age population is inapplicable here because, unlike the circumstances found in this case, the Solomon Court was considering the use of black registered voting population for the first Gingles factor, which tests whether a minority group is large enough to be able to elect a candidate from a single-member district. See Solomon, 899 F.2d at 1018 (Kravitch, J., concurring); see also Gingles, 478 U.S. at 50-51 n. 17, 106 S.Ct. at 2766 n. 17. The Solomon Court evaluated a population where blacks had been consistently frustrated in their ability to elect their chosen representatives — a situation that would tend to discourage blacks from registering to vote. See Solomon, 899 F.2d at 1018 (Kravitch, J., concurring) ("An at-large election system that frustrates the ability of minorities to elect their chosen representatives will naturally reduce the incentive for blacks to register to vote."); see also Negron v. City of Miami Beach, 113 F.3d 1563, 1568 (11th Cir.1997) (explaining the Solomon case). Use of BREG population rather than BVAP would underestimate the ability of blacks to form a group sufficiently large and geographically compact enough to elect its own representative. Therefore, the Solomon Court rejected the use of an indicator "which might reward and perpetuate a history of disenfranchising blacks." Solomon, 899 F.2d at 1018 (Kravitch, J., concurring). Plainly, this concern is not an issue in the case before us.

The defendant also incorrectly relies on the Supreme Court's decision in De Grandy, 512 U.S. 997, 114 S.Ct. 2647. In De Grandy, the Court used voting-age population to determine whether "the percentage of single-member districts in which minority voters form an effective majority mirrors the minority voters' percentage of the relevant population." Id. at 1000, 1017, 114 S.Ct. at 2651, 2660. However, the Court expressly declined to resolve the issue of whether the "relevant" population is "the minority group's share of the population, or of some subset of the population," since the issue was not dispositive in that case. Id. at 1018 n. 14, 114 S.Ct. at 2660 n. 14. Moreover, the Court made no finding as to the appropriate measurement for determining whether a minority constitutes a sufficient population for a district to be considered a majority-minority district. Indeed, limiting the Court's § 2 analysis to consideration of voting age population figures would contradict De Grandy's general holding that "[n]o single statistic provides courts with a shortcut to determine whether a set of single-member districts unlawfully dilutes minority voting strength." Id. at 1020-21, 114 S.Ct. at 2661-62. Accordingly, in Georgia v. Ashcroft, the district court decided to "evaluate the entire picture, taking into account the different measurements of eligible African American voting population." 195 F.Supp.2d 25, 79 (D.D.C.2002) (three-judge court). On appeal, the Supreme Court also considered both BREG and BVAP. Ashcroft, 539 U.S. 461, ___ n. 1, 123 S.Ct. 2498, 2508 n.1.

Likewise, in this case, the Special Master considered both registered voter population and voting age population. See Persily Aff. at 15 ("We concentrate here on the data concerning the 2002 African-American registered voter population (BREG), but similar data broken down by African-American voting age population (BVAP) are presented in Exhibit Q. The distribution of the African-American population among districts as revealed by the two statistics is comparable, and we paid attention to both statistics as we constructed our plans."). The Special Master noted that blacks make up 26.72% of Georgia's voting age population and 25.62% of registered voters. In the Senate there are thirteen majority-minority districts, or 23.20% of the total, using either of the measures. In the House, African-Americans are a majority of the registered population in forty-four districts (24.44%) and a majority of the voting age population in thirty-nine districts (21.67%). We can discern very little difference in the relevant numbers. Here, the Special Master considered both BREG and BVAP and ensured that there was no dilution or retrogression using either set of figures.

The defendant's second argument is equally unpersuasive since, using the defendant's own BVAP figures, the number of districts with more than 25% BVAP actually increases from the Senate Benchmark Plan and remains constant in the House, as compared to the House Benchmark Plan. Using BVAP, there were twenty-five districts in the Senate Benchmark Plan with greater than 25% of the electorate, as compared to twenty-seven such districts in the Special Master's 1-B Senate Plan. Both the House Benchmark Plan and the Special Master's 1-B House Plan had eighty such districts. Moreover, the 1-B House Plan has thirty-nine majority-minority districts and forty-one influence districts, while the Benchmark House Plan had only thirty-seven majority-minority districts and forty-three influence districts.

The defendant suggests that we are required to defer to the choice made by the General Assembly during the 2001-2002 redistricting process to increase black influence districts while maintaining the number of black majority-minority districts and that this Court may not offset reductions in influence districts with increases in the number of majority-minority districts. However, the Enjoined Plans cannot serve as a point of comparison for the Court's retrogression analysis. As we noted in our February 10, 2004 Order, one of the problems with the underlying maps was the state's systematic underpopulation of districts in and around urban Atlanta — a procedure that allowed it to create a greater number of districts with high black population percentages. Larios, 300 F.Supp.2d at 1342-47. Most of the majority-minority districts in the Enjoined Plans were discernibly underpopulated, and the districts in the 30% to 49% range under the Enjoined Plans were some of the most underpopulated in the state. In short, the proper comparison is to the Benchmark Plans, not the Enjoined Plans.

The defendant also points to the 1-B House Plan's decrease in the number of influence districts as a source of retrogression. The defendant's analysis, however, fails to examine the plan as a whole. In Georgia v. Ashcroft, the Supreme Court held:

In examining whether the new plan is retrogressive, the inquiry must encompass the entire statewide plan as a whole .... Thus, while the diminution of a minority group's effective exercise of the electoral franchise in one or two districts may be sufficient to show a violation of § 5, it is only sufficient if the covered jurisdiction cannot show that the gains in the plan as a whole offset the loss in a particular district.

539 U.S. at ___, 123 S.Ct. at 2511. Moreover, "any assessment of the retrogression of a minority group's effective exercise of the electoral franchise depends on an examination of all the relevant circumstances .... No single statistic provides courts with a shortcut to determine whether a voting change retrogresses from the benchmark." Id. (internal citations and quotation marks omitted). In the 1-B House Plan, the loss of two influence districts, as compared to the House Benchmark plan, is offset by the gain of two majority-minority districts. In short, we agree fully with the Special Master that there is neither dilution nor retrogression in either of the plans adopted by this Court.

Finally, we conclude that the Special Master's 1-B Plans are faithful to Georgia's traditional redistricting principles of compactness, contiguity, minimizing the splits of counties, municipalities, and precincts, recognizing communities of interest,^[18] and avoiding multi-member districts.

First, compared with both the Enjoined Plans and the Benchmark Plans, the Special Master's 1-B Plans equal or exceed the prior plans' compactness, whether using smallest-circle or perimeter-to-area compactness measures.^[19] Under the smallest-circle measure, the 1-B Senate Plan's .42 level equals the level attained by the Senate Benchmark Plan and exceeds the .35 level of the Enjoined Senate Plan. Under the perimeter-to-area measure, the 1-B Senate Plan's .28 level matches the level attained in the Senate Benchmark Plan and substantially exceeds the .16 level of the Enjoined Senate Plan. Likewise, the 1-B House Plan's smallest-circle measure of .41 level matches the level attained in the House Benchmark Plan and exceeds the .38 level of the Enjoined House Plan. Under the perimeter-to-area measure, the 1-B House Plan's .30 level exceeds the .29 level attained in the House Benchmark Plan and the .24 level of the Enjoined House Plan. In addition, the Special Master's Plans achieve these compactness scores along with population deviations fully below ±1%, in sharp contrast to the substantially higher population deviations found in the four comparison plans.

Similarly, the degree of contiguity reflected in the Special Master's Senate and House Plans contrasts sharply with the Enjoined Plans. We found that with respect to the Enjoined Plans:

[D]istrict contiguity was not a real concern among plan drafters and legislators.... While all of the districts are technically contiguous (as required by state law), many districts achieve that designation through the use of water contiguity, which is predicated on the assumption of line-of-sight across a lake or other body of water, or touch-point contiguity, which is predicated on facing corners in a checker-board like fashion.

Larios, 300 F.Supp.2d at 1332. Unlike the Enjoined Plans, the districts in the Special Master's 1-B Plans are fully contiguous, except where contiguity is impossible.^[20] Moreover, none of the districts in the Special Master's 1-B Plans exhibit touch-point contiguity, except where such contiguity is an existing feature of a county.^[21]

Compactness and contiguity were naturally enhanced by the use of counties and municipalities as building blocks wherever possible. "Georgia has an unusually high number of counties: 159, the greatest number of any State in the Union apart from the much-larger Texas. These small counties represent communities of interest to a much greater degree than is common...." Abrams, 521 U.S. at 100, 117 S.Ct. at 1940. In addition, avoiding county splits is also important because

[e]ach county, municipality, or other jurisdiction has a local delegation and any legislator whose district encompasses territory within a specific city or county is a member of the local delegation for that entity.

The local delegations make recommendations to the House and Senate standing committees, which then recommend local legislation to the entire body. A local bill must receive the requisite majority from the local delegation to be reported favorably out of the standing committees with a "do pass" recommendation.

DeJulio, 290 F.3d at 1293 (footnote omitted). Furthermore, "[i]f local legislation has received the requisite number of signatures of representatives or senators whose districts lie partially or wholly within the locality which the legislation affects, it is ordinarily passed on an uncontested basis as a matter of local courtesy." Id. at 1293-94. Thus, having a district intrude across county (or municipality) lines gives a legislator whose district predominately lies outside that county (or municipality) a vote on issues that may well not directly affect the majority of the legislator's constituents.

The 1-B Senate Plan keeps 119 counties entirely within a single district and splits forty counties into 100 parts, thirty-four of which are only divided between two districts. Of the remainder, two counties are each divided among three districts; one county is divided among five districts; and three counties are each divided among seven districts. The majority of such divisions are an inevitable result of the population concentration in the Atlanta metropolitan area. This too provides a sharp contrast with the Enjoined Senate Plan, where eighty-one counties were split into 219 parts.

Although the lower ideal population size for a House district as compared to the ideal population size of a Senate district necessarily results in a greater number of county splits, the 1-B House Plan also substantially reduces the number of county splits. In contrast to the Enjoined House Plan, which split eighty of the state's 159 counties into 266 parts, the 1-B House Plan keeps eighty-two counties entirely within a district and splits seventy-seven counties into 278 parts, of which fifty-nine are split either between two districts or among three districts. Of the remainder, seven counties are each divided among four districts; four counties are each divided among five districts; two counties are each divided among six districts; and the remaining five counties are divided among seven or more districts. As with the county splits in the 1-B Senate Plan, the majority of such divisions are an inevitable result of the population concentration in the Atlanta metropolitan area.

The comparative numbers, however, do not tell the full story. In fact, the difference is far greater than it appears because the 1-B House Plan — which uses no multi-member districts — has thirty-three more districts than the Enjoined House Plan and fifty-six smaller districts, as a result of dividing larger multi-member districts into smaller single-member districts. The Enjoined House Plan had 147 districts, with 124 single-member districts, fifteen two-member districts, six three-member districts and two four-member districts.

Moreover, the lower number of county splits in both of the Special Master's 1-B Plans was achieved under the primary constraint of achieving population equality. Relaxing that constraint would inevitably have reduced the number of splits. Accordingly, since the Special Master's districts achieve population deviations of below ±1% — whereas the Enjoined Plans' deviations were approximately ±5% — the reduction of county splits in the Special Master's 1-B Plans is all the more significant.

The Special Master also attempted to adhere to the principle of keeping municipalities contained within districts. This proved particularly difficult, as many municipalities cross county lines. Thus, Braselton (on the border of Gwinnett and Barrow counties), Maysville (on the border of Banks and Jackson counties), Waycross (on the border of Ware and Pierce counties) and Villa Rica (on the border of Carroll and Douglas counties) cross county boundaries. Even more complicating was the fact that the annexation patterns of a number of municipalities have resulted in non-contiguous municipalities (Byron, Calhoun, Cartersville, LaGrange and Warner Robins have non-contiguous portions).

The Special Master also considered at length the defendant's specific objections to the splitting of certain precincts in the Special Master's Original Plans. To the extent possible, and without degrading principles of greater importance or causing partisan effects, the Special Master adopted any changes that were unopposed and did not result in splits of cities made whole (or nearly whole) under the Special Master's 1-B Plans. In some cases, changes that by themselves were acceptable were not adopted because they required still other changes that did not meet the basic criteria.^[22]

The Special Master's 1-B Plans also followed this Court's direction that "absent insurmountable difficulties," the court-drawn plans should not use multi-member districts. Connor v. Johnson, 402 U.S. at 692, 91 S.Ct. at 1762. The 1-B House Plan removed all of the multi-member districts, adding a total of thirty-three single-member districts.

Finally, we are fully satisfied that the Special Master's 1-B Plans adhere to the strict guidelines given to the Special Master regarding incumbency protection. In a strictly neutral and consistent manner, the Special Master unpaired forty-five of the incumbents who had been placed in the same districts by the Special Master's Original Plans.^[23] While the Special Master's Original Plans contained a total of seventy-three paired representatives,^[24] the Special Master's 1-B Plans paired only twenty-eight representatives. In making recommendations, the Special Master rejected any change if it would cause a departure from the primary considerations of compliance with the one person, one vote principle and the Voting Rights Act and adherence to other traditional redistricting principles.^[25] Furthermore, the Special Master did not recommend changes where the requested change either would cause problems in adjoining districts or require substantial changes to the Special Master's Plans.^[26]

In short, we conclude that these plans follow the constitutional mandate of one person, one vote, comply fully with § 2 and § 5 of the Voting Rights Act, and carefully adhere to Georgia's traditional principles of redistricting. Accordingly, we hereby adopt the Special Master's 1-B Plans as the Court's plans and again direct the Secretary of State and Commissioner of Elections to implement these plans promptly and consistently with the Constitution and laws of the United States and the Constitution and laws of the State of Georgia.

NOTES

[1] In Bonner v. City of Prichard, 661 F.2d 1206 (11th Cir.1981) (en banc), the Eleventh Circuit adopted as binding precedent all decisions of the former Fifth Circuit rendered prior to October 1981.

[2] The Special Master and his experts received substantial assistance from personnel at the Georgia Legislative Reapportionment Office pursuant to court order. All of the Special Master's redistricting plans were developed using computers at the Legislative Reapportionment Office. The primary computer program was "Maptitude for Redistricting" ("Maptitude"), widely used for legislative redistricting throughout the country and currently employed for redistricting purposes by the General Assembly and the Legislative Reapportionment Office.

[3] The parties and amici curiae filed the following comments and objections on March 19:(i) Plaintiffs' Comments and Corresponding Proposals regarding Report and Recommendation of Special Master; (ii) Defendant's Objections to Report and Recommendation of the Special Master ("Defendant's Objections"); (iii) Brief Amicus Curiae and Comment — Map Proposal of Georgia Legislative Black Caucus ("GLBC Comments"); (iv) Proposed Comments of Rep. Thomas B. Buck, III; (v) Comments and Objections of Board of Commissioners of Talbot County, Georgia to Special Master's Proposed Reapportionment ("Talbot County Objections"); and (vi) Comments and Objections of the Democratic Party Intervenors on Special Master's Report and Recommendation, filed by Robert S. Kahn and the Democratic Party of Georgia ("Democratic Party Comments"). On March 22, 2004, the Democratic Party filed Rebuttal Comments and Objections of Democratic Party Intervenors on Special Master's Report and Recommendation.

[4] For this initial review after receiving the March 19 Comments, the Special Master considered political incumbency only where a party or an amicus specifically requested a change concerning an incumbent pairing. We later directed the Special Master to consider all incumbent pairings, regardless of whether they were specifically requested or raised in comments.

[5] The Board of Commissioners of Talbot County, Georgia and Representative Buck submitted written comments and objections on March 19, 2004, but were not represented at the March 22, 2004 hearing. The Democratic Caucus appeared at the March 22, 2004 hearing but did not file a written comment or objection on March 19, 2004.

[6] Comments and objections lodged on March 23, 2004 included the following: (i) Plaintiffs Comments and Corresponding Proposals Regarding Supplemental Report and Recommendation of The Special Master and Notice to Court of Incumbents Who Have Announced Intention Not to Seek Reelection to the Georgia General Assembly; (ii) Comments and Objections of Democratic Party Intervenors on Special Master's Supplemental Plan of March 22, 2003; (iii) Amicus Curiae Brief of Democratic Caucus of House of Representatives; (iv) Declaration of Robert S. Kahn; and (v) compilation of objections expressed by Senator Ralph T. Hudgens and the Madison County Chamber of Commerce to the splitting of Madison County. On March 24, 2004, two further comments were received and reviewed by this Court: (i) Stipulation of the Democratic Party of Georgia and (ii) Supplement to Amicus Curiae Brief of Democratic Caucus of House of Representatives.

[7] The same parties and amici curiae participated in the March 25, 2004 hearing as did in the March 22, 2004 hearing. Two comments were filed in open court: (i) Comments of the City of Lagrange to Special Master's Report and Recommendation 1-B; and (ii) Plaintiffs' Response in Opposition to Motion to Intervene of the Democratic Party of Georgia and its Chair. Plaintiff Senator Charles B. Tanksley's Comments Concerning District 32 was also filed in the clerk's office on the same day.

[8] At a hearing on March 23, 2004, the Secretary of State asked this Court to adopt two procedural modifications, one modifying time requirements for processing absentee ballots and the other modifying time requirements for notifying voters of the district changes. We adopted both suggestions in an Order entered on March 30, 2004.

[9] The high and low deviations do not add up to the total deviation figure due to rounding.

[10] As with the 1-B Senate Plan, the high and low deviations do not add up to the total deviation figure due to rounding.

[11] We recognize that even when "the percentage of single-member districts in which minority voters form an effective majority mirrors the minority voters' percentage of the relevant population," De Grandy, 512 U.S. at 1017, 114 S.Ct. at 2660, such proportionality does not provide a safe harbor against a § 2 challenge. Id. at 1017-21, 114 S.Ct. at 2660-62. Proportionality is, however, "always" relevant evidence of the lack of dilution that a court must consider. Id. at 1025, 114 S.Ct. at 2664 (O'Connor, J. concurring). Outside of the objections raised by the defendant, which we discuss later, infra pp. 1366-1369, no party or amicus curiae suggests that the Special Master's 1-B Plans illegally constrict any minority group's access to the political process, nor has any party or amicus curiae made a specific § 2 challenge as to any particular area of the state. By contrast, both the Democratic Party and GBLC say that the plans fully comply with § 2. In the absence of any evidence that the plans dilute minority voting strength, the number of majority-minority districts, as compared to the percentage of African-Americans in the population as a whole, provides relevant evidence of compliance with the racial-fairness mandates of § 2.

[12] The Benchmark Plans were used for purposes of the § 5 analysis under the Voting Rights Act. See Abrams, 521 U.S. at 97, 117 S.Ct. at 1939 ("Nor can the 1992 plan, constitutional defects and all, be the benchmark. Section 5 cannot be used to freeze in place the very aspects of a plan found unconstitutional."). However, in comparison to both the Benchmark Plans and the Enjoined Plans, the Special Master's 1-B Plans meet or exceed the requirements of § 2 and § 5 of the Voting Rights Act.

[13] The number of majority-minority districts in the Senate was thirteen, considering either the proportion of black registered voters or the black voting age population. The Enjoined Senate Plan also had thirteen majority-minority districts.

[14] The Enjoined Senate Plan had one more Senate district with 30% to 49% African-American Registered Voters, but one less influence district with less than 30% African-American Registered Voters.

[15] These figures use black registered voters to measure the proportion of blacks. Under the defendant's black voting age population figures, there were thirty-nine majority-minority districts in the Special Master's 1-B House Plan and thirty-seven majority-minority districts in the Benchmark House Plan.

[16] See March 22, 2004 Hearing Tr., at 69-72, 74, 77-78. Specifically, the Court and Mr. McDonald, representing the GBLC, had the following exchange:

COURT: It is your view that the plan neither retrogresses — I'm talking about with the amendation — under §5 — nor dilutes under § 2?

MR. MCDONALD: That is correct, Your Honor.

Id. at 71. Similarly, Mr. Bramlett, representing the Democratic Party, stated, "We agree that with the changes that deal with African-American legislators, that the Court has now taken out the voting rights objections that we have to the Special Master's original plan." Id. at 74.

To the extent claims initially were made that the plans retrogressed because black incumbents in the Special Master's Original Plans were paired more frequently than white incumbents, the final Special Master's 1-B Plans, adopted by this Court, unpaired incumbents in a neutral manner and ultimately paired only four black incumbents and twenty-four white incumbents. The Special Master's 1-B Plans unpaired the incumbents in a wholly neutral and impartial way, without concern to race or politics.

[17] Then Chief Judge Tjoflat's concurrence adheres to the views expressed by the original three-judge court, of which he was a member. Although not specifically mentioned in his en banc concurrence, he rejected the district court's use of black voting age population in the original panel opinion. See Solomon v. Liberty County, 865 F.2d 1566, 1574 (11th Cir.1988), vacated, 873 F.2d 248 (11th Cir.1989) (en banc).

[18] We did not direct the Special Master to preserve the cores of districts in the invalidated plans. In the Enjoined Plans, "[t]o the extent that the cores of prior districts were preserved at all, it was done in a thoroughly disparate and partisan manner .... Quite simply, the population deviations in the House Plan and the 2002 Senate Plan did not result from a neutral, consistently applied concern for retaining incumbent cores." Larios, 300 F.Supp.2d. at 1350-51.

Because the Enjoined Plans did not evince an intent to preserve the cores of prior districts, the question arises about what "cores" exist to be preserved. However, to the extent possible, and within the constraints of the guidelines, the Special Master's Senate and House 1-B Plans were drawn with recognition of the cores of districts shown in either the Enjoined Plans or the Benchmark Plans, under the theory that the prior districts should reflect communities of interest.

[19] The "smallest-circle" measure compares a district's area to that of the smallest circle that could encompass the entire district; the "perimeter-to-area" measure is the ratio of the district's area to the area of a circle with the same perimeter. Under both measures, the figure for the "perfect" district would be 1.00.

[20] For example, the districts encompassing the coastal counties of Eastern Chatham, Bryan, Liberty, McIntosh, Glynn and Camden cannot be fully contiguous because portions of those districts are islands.

[21] Portions of Lee, Peach and Rockdale counties are touch-point contiguous with the main bodies of those counties; therefore, the districts encompassing those counties similarly may reflect touch-point contiguity.

[22] For example, the defendant proposed moving precincts in District 44, but this change was not adopted because of objections raised as to Voting Rights Act concerns. Likewise, the Special Master accommodated only one of the defendant's proposed changes to District 10, since others required population shifts that would either split cities or take districts out of ideal deviation range.

[23] The defendant objected to the Special Master's initial consideration of only those incumbent pairings which were specifically objected to in the March 19 Comments. The defendant argued that other parties and amici curiae should have the opportunity to express their objections and suggestions concerning incumbent pairings and that the pairings should be considered without regard to whether they had been specifically proposed or objected to. We agreed, and accordingly granted all interested parties and amici curiae additional time to submit further objections and suggestions regarding incumbent pairings and directed the Special Master to consider all possible incumbent pairings, without regard to whether a party or amicus curiae had objected to the pairing. See supra note 4.

[24] The Special Master's 1-B Report and Recommendation indicates that the Special Master's Original Plans paired eighty-seven incumbents, while the Special Master's 1-B Plans paired only twenty-eight incumbents. However, after taking into account incumbents who expressed their intent not to run for re-election in the Georgia General Assembly, the number of paired incumbents in the Special Master's Original Plans was reduced to seventy-three.

[25] For example, the Special Master rejected the GBLC's suggestion to change Districts 49 and 51 to resolve an incumbent pairing, because the proposed solutions would have created a noncompact district.

To the extent that parties and other interested individuals and entities advocated changes other than the unpairing of incumbents, those changes were rejected out of concern for the partisan effect the changes might have. Requests to unite communities of interest, changes to repair county or city splits, or incumbency-related interests unrelated to pairings were rejected because the Special Master was justifiably concerned that the particular requests were made largely for partisan reasons. Nevertheless, as noted, some suggestions concerning precinct splits were accepted, since they adhered to the neutral districting principle of avoiding precinct splits, but only if they were unopposed and no superseding principle was degraded.

One non-partisan change that was adopted by this Court was a modification to the numbering system, suggested by both the defendant and the plaintiffs in their March 19 Comments. In the Order entered on March 25, 2004, we directed that the districts set forth in the 1-B Plans be renumbered in accordance with the proposal of the Secretary of State, adopted by the Order as Attachment A. It should be noted that throughout this Memorandum Opinion, when discussing proposed changes to the Special Master's plans, we have referred to districts according to the numbering system used in the Special Master's plans prior to March 25, 2004.

[26] Thus, for example, the Special Master rejected the Democratic Party's suggestion of moving Senator Tolleson from District 20 into District 18, moving Senator Gillis from District 4 to District 20, and moving parts of District 14 into District 18, because the proposal would have involved major changes to the map that span nearly the entire state.