Georgia v. Ashcroft

support employment. standings Absent such claims of entitle- right to federal .that Roth, ment to those the court declines to extend benefits.” 408 U.S. at precedent, (holding respon- 92 S.Ct. 2701 that the prece- for which there is no rights those property dent did not have a interest —the Through alleged their denial of a dent. component substantive employ- interest in federal re- fundamental —sufficient quire a ment, hearing procedural compo- plaintiffs attempt to trigger —the government nent —when the declined re- Pis.’ scrutiny analysis. strict See Mem. newal employment). of his contract of Mot. for Supp. of Summ.J. 89. plaintiffs Since have not raised a denial defendants, however, oppose plaintiffs’ procedural process, due the court need scrutiny that strict argument applies to not raise that sponte. issue sua There- Supp. Mem. in given facts. See Defs.’ fore, the court summary judgment enters at 10. The Mot. for Summ.J. defendants on Count Two for the defendants and the argue also there is no fundamental intervenor-defendants, and denies sum- right employment to federal that would mary judgment plaintiffs as to the plaintiffs’ entitle the loss of federal em- Count Two. ployment process pro- to substantive due id. The tection. See intervenor-defen- IV. CONCLUSION concur with the dants defendants on this reasons, For the foregoing the court point. Supp. See Int.-Defs.’ Mem. grants the defendants’ motion for sum- agrees Mot. for Summ.J. at 37. The court mary judgment grants the intervenor- is, position with the defendants’ since there summary defendants’ motion for judgment indeed, recognizes no case law that a fun- on both Accordingly, Counts One and Two. right employment, damental to federal plaintiffs’ the court denies the motion for procedural process guarantee. due summary judgment on both Counts One Roth, See 408 U.S. at 92 S.Ct. 2701. directing parties and Two. An order process analysis Therefore substantive due a manner consistent with this Memoran- apply not here. does See id. Opinion separately contempo- dum Nonetheless, previ alluded to raneously day issued this 29 of March ously, component there is a second to due process analysis, otherwise referred to as procedural process. due See id. When an

underlying subject law is to substantive process analysis

due and survives strict

scrutiny, it implemented must also be such, fair manner. See id. As the second GEORGIA, Plaintiff, The State of proscription under the Due Process Clause v. of both the Fifth and Fourteenth Amend ASHCROFT, al., et John Defendants. any proce ments is measure denies 01-2111(EGS)H. process. dural due See Mathews v. El No. Civ.A. 319, 335, dridge, U.S. 96 S.Ct. Court, United States District (1976). L.Ed.2d Property interests are District of Columbia. Constitution, generated by “Math April they er are created and their dimensions by existing are defined rules under

standings independent that stem from an

source such as state law rules or under- McDonald, Nojeim,

Gregory Laughlin T. Bell, E.B. American Civil Liber- Meredith Inc., Atlanta, Foundation Geor- ties Union gia, for movant. EDWARDS, Judge, Circuit

BEFORE: SULLIVAN, Judge, District OBERDORFER, Judge. District Senior by District Opinion for the court filed SULLIVAN, Judge Judge which Circuit joins, and in HARRY T. EDWARDS Judge which Senior General, joins Parts III.C.2. Ashcroft, Attorney OBERDORFER D. John Jr., Concurring opinion filed Attorney and III.C.3. Assistant Ralph Boyd, F. EDWARDS, in Division, Judge HARRY T. General, Joseph D. Circuit Rights Civil Becker, joins. Judge SULLIVAN which District Rich, Rengle, David J. Robert A. Gear, part dissenting Tucker, Opinion concurring Bruce I. Thomas James Justice, Judge part by filed Senior District Walsh, Department D. James Division, OBERDORFER. Rights Voting Section/Civil D.C., for defendants. Washington, SULLIVAN, Judge. Baker, Attorney General of E. Thurbert declaratory judg- This is an action for Dunn, Dennis R. Georgia,

the State ment commenced the State General, Attorney State Senior Assistant Act Voting Rights 5 of the under Section Atlanta, GA, Mark H. Department, Law (1994) (“Section 1965, § 42 1973c U.S.C. Cohen, Attorney Gener- Special Assistant 5”). declaratory judg- seeks a The State L.L.P., Atlanta, GA, Sanders, al, Troutman plans passed by redistricting that the ment Walbert, Attor- Special Assistant F. David the Assembly for Georgia General Chesin, General, Parks, & ney Walbert and the Congressional seats United States Atlanta, GA, Samp- Miller, P.C., Thomas do not “have and House seats State Senate son, Sr., Attorney Special Assistant Gener- the effect of and will not have purpose al, Thomas, Sampson & Patter- Kennedy, to abridging right vote denying or Pierson, son, Atlanta, GA, Fries Stuart membership color” or account of race or General, Attorney Wash- Special Assistant 42 U.S.C. group. language D.C., plaintiff. ington, for § 1973c. Braden, Jr., Ellis, Bak- E. Mark Lee T. imposes weighty Rights Act Voting The D.C., L.L.P., Washington, & Hostetler er history with a jurisdictions obligations on Lewis, Strickland, Anne W. Frank B. in their electoral of racial discrimination L.L.P., At- Brockington Lewis Strickland Act enacted the Congress processes.1 lanta, for intervenor-defendants. Georgia, country “firm to rid intent[ ] with voting” by a Jonesboro, GA, of racial discrimination pro se King, B. Michael stringent remedies.” “complex scheme of movant. eligibility determine device was used to test or jurisdictions are determined 1. Covered 1973b(b) Georgia is a covered Act, year. Id. vote in that § to 42 U.S.C. 4 of the Section 5, (Mar. App. § 28 C.F.R. (2001), partic- jurisdiction. where voter states and include 1964, and where a ipation was below 50% 2002). 30 Katzenbach, v.

South Carolina prerequisite U.S. cation or voting, or stan- 301, 315, 803, 812, dard, 86 S.Ct. practice, 15 L.Ed.2d 769 procedure with respect (1966). The Georgia General Assembly voting.”3 § 42 U.S.C. 1973c. Two ave- aware of statutory well its preclearance constitu- nues for provided by are responsibilities, tional has Act. Id. The covered as the State jurisdiction may seek spent much of the last defending decade a declaratory judgment a three-judge legislative reapportionment plans its District Court for the District of Columbia against claims of racial gerrymandering, practice the new does not have the brought pursuant 2 of purpose Section the Vot- or effect of denying or abridging ing Rights Act and the United right States Con- to vote on account of race or See, Johnson, e.g., Abrams v. stitution. Id. alternative, color. jurisdic- 521 U.S. 117 S.Ct. 138 L.Ed.2d may tion submit proposed procedures its (1997) (affirming court-ordered Attorney Con- to the approval; General for gressional Miller v. redistricting plan); procedures if, approved deemed after Johnson, 515 U.S. days S.Ct. following filing a completed (1995) L.Ed.2d 762 (holding submission, that Congres- Attorney General has not sional redistricting plan equal pro- violated any objections raised to the pro- clause). tection Id. cedures. obligations State’s under Section Supreme Court characterized

however, differ significantly from those un- unusual, Section 5 as “an and in some der Section 2 of the Voting Rights severe, Act.2 aspects a procedure insuring Section requires specific jurisdictions to. that states would not discriminate on the comply “preclearance” procedures be- basis of race the enforcement of their implementing fore any “voting new qualifi- voting laws.” Allen v. State Board of Voting Rights Section of the provides Act person right shall be denied the to vote for voting that qualification prerequisite "[n]o or comply failure to qualification, with such standard, voting, to practice, or procedure or standard, prerequisite, practiсe, proce- or imposed shall be applied by any or State or Provided, qualification, dure: that such political a subdivision in manner which re- standard, prerequisite, practice, proce- or abridgement sults in a denial or right of may dure pro- enforced without such any citizen of the United States to vote on ceeding qualification, if the prerequisite, account of race or color.” 42 U.S.C. standard, practice, procedure or has been 1973(a). § by legal submitted the chief officer or other appropriate of official such State or subdivi- provides: 3. Section 5 sion Attorney to the General and the Attor- political Whenever a State or subdivision ney interposed General objection has not an respect with prohibitions to the which set submission, sixty days within after such ex- 4(a) forth in section are in effect shall enact cept that Attorney neither the or General’s seek to fail- any voting qualifica- administer object ure prerequisite standard, declaratory tion to a voting, judgment or nor to or practice, procedure respect or entered under this section shall vot- bar a subse- ing different quent from enjoin that force or effect on action to enforcement of such 1, 1964, November State or qualification, standard, such subdivi- prerequisite, prac- may sion institute tice, an action in the United procedure. Any or action under this States District Court for the District of Co- section shall be heard and determined lumbia declaratory judgment for a that judges court of three in accordance with standard, qualification, such prerequisite, provisions the of section 2284 of title 28 of practice, procedure or does not have the the any appeal United States Code and shall purpose and will have not the effect of Supreme lie to the Court. denying right or abridging the to vote on (1994). § 42 U.S.C. 1973c color, account of race or and unless and until the court judgment such enters no that 817, declaratory judgment United 544, 556, Elections, 89 S.Ct. 393 U.S. plan, (1969). Congressional redistricting intend- States 5 was Section 22 L.Ed.2d 2EX11, the State House redis- rapid mecha- Act No. provide an efficient ed to 2EX23, pro- satisfy the tricting plan, Act No. changes preclearing for nism hold, that how- cedures, expressly providing of 5. We requirements while Section ever, way Georgia in no affects of has not met preclearance that the State such plan challenge 5 with ability proof of individuals its burden of under Section 549, 556, 89 S.Ct. redistricting Id. at grounds. other to the State Senate regard has not demonstrated plan. State of the evidence that preponderance prohibits States Section redistricting plan does not State African opportunities of diminishing the purpose and will have their electoral voters to exercise right denying abridging effect States, 425 U.S. Beer v. United power. Accord- on account of race or color. vote 130, 141, 47 L.Ed.2d 96 S.Ct. declaratory for a request ingly, State’s (1976). has demonstrated that the meets judgment State increasingly African American voters 5 is denied. requirements Section heard to make their voices have been able indicates, record howev the ballot. The History Preliminary I. Procedural er, within the State that there are areas Matters persists. polarized voting racially where areas, consistently In these white case, foregone In this State preferred candidates against vote applying Attorney to the General option of and district Americans local African redistricting plans, its preclearance elections, African Amer strength so the in this court. Section and has filed suit part on the rests substantial ican votes existing districting essentially freezes the African American voters numbers of sheer and until a declara- plans Georgia unless there is evidence in a district. Where from this court tory judgment is obtained voting, redistricting racially polarized plans proposed reapportionment votes African American plan that reduces discriminatory purpose or ef- are without offsetting gains else in a district with no Bd., v. Parish School fect. Reno Bossier impermissible specter raises the where 1491, 471, 477, 137 117 S.Ct. 520 U.S. situation, the In this State retrogression. I”). (“Bossier (1997) L.Ed.2d 730 demonstrate that there hard-pressed to 10, suit on October Georgia filed State in African “backsliding” no has been 2001, a the court enter requesting Reno v. Bossi voting strength. American Congres- declaratory judgment 320, 335, Bd., 528 U.S. er Parish School plans sional, House and State State (“Bossier 866, L.Ed.2d 845 120 145 S.Ct. discriminatory purpose or have a do not II”). in a a failure is fatal And such primary is Georgia’s general effect. case, the burden is on 5 because Section 16, 2002, and the State July for scheduled redistricting the State to show allow recently preclearance received to opportu adversely affect the plan will not leg- and the state Congress candidates to effec African American voters nities of primary for the qualify islature to franchise. their electoral tively exercise 21, Opp’n 2002. See Defs.’ 10 to June June 141, Beer, 96 S.Ct. 1357. 425 U.S. (discuss- Expedited Trial at to Mot. for 26, preclearance November ing the evidence carefully reviewing After (2001)). light 2EX us, Act hereby grant a we the record before “extraordinary” remedy by mandated putative interests of the intervenors. Act, 24(a). Voting Rights the court has acted See Fed.R.Civ.P. possible speed with all expeditiously 31, 2001, On December the United Allen, resolve this matter. 393 U.S. at position States identified its respect with 563, 89 S.Ct. 817. to the redistricting plans. On scheduling 4, 2002, court’s January order set a de- Jones filed a renewed schedule, manding briefing permit- while motion to receiving intervene. After ting parties engage response motion, extensive and reply to this on Jan- discovery up 10, until 2002, the commencement of uary the court granted Jones’ Order, the trial. See Civ. Action No. 01- motion to intervene required and the inter- 20, Indeed, Dec. 2001. with the con- comply venors to with the Court’s initial parties, sent of the Order, commencement of the scheduling pretrial order. See trial was days deferred for three 01-2111, to enable Action Civil No. Jan. parties complete discovery. (granting intervention as to State House plans); Order, and State Senate Civil Ac- At the time of the initial scheduling con- 01-2111, tion No. (granting Jan. ference, the United yet States had not Congressional intervention as to redistrict- identified position respect its to each plans). below, ing As discussed Mr. King plans. of the submitted Upon consider- renewed his motion to intervene in an ation by of motion Georgia, the State of fashion, untimely and was denied leave to response argument thereto and oral at the intervene. conference, scheduling the court required the United identify States to legal posi- its With the consent parties, Judge by tion no later than December presided Sullivan four-day over the trial.4 Following trial, the conclusion of the Two motions to intervene were filed ear- parties submitted findings of fact ly in the proceedings, by one four African law, post-trial conclusions of mem- Georgia, citizens of Patrick oranda of Closing arguments law. were Jones, Tyra, Roielle Della Steele and heard the three-judge panel on Febru- (“Jones”), Georgia Benton one Mi- ary chael King, an African American lawyer and resident of Senate District 44. Both There are several preliminary matters motions prejudice were denied without fol- that the court must address before focus- lowing the court’s order that the ing United reapportionment on the three plans. States identify Order, its legal position. (1) Pending before the court are: a motion 01-2111, Civil Action 2001; No. Dec. for leave to file an amicus curiae brief Order, 01-2111, Civil Action No. Dec. by submitted the American Civil Liberties 2001. The court (2) invited the (“ACLU”); movants to Union plaintiffs motion to briefs, file that, amicus curiae but held strike the Jones intervenors for lack of without (3) clarification of the United standing; States’ King’s Mr. stay motion to legal position, it could not determine if the proceedings and motion for reconsidera- existing parties adequately represented intervene; (4) tion of his motion to 4. The testimony bulk of in this case deposition was sub- nate testimony place in of cross- paper, mitted on in accordance with the examining the witnesses at trial. The live scheduling court’s order and with the consent trial, therefore, testimony at consisted of parties. While the court’s order direct- questioning cross-examination redirect ed testimony that direct be submitted in parties’ expert witnesses. writing, parties permitted desig- were to Nevertheless, reside whether intervenors motions to ex- intervenors’ defendants’ expert testimo- or districts plaintiffs benchmark portions of clude not affect their in this matter does ny. issue standing purposes challenging for Leave to Partici- Motion for A. ACLU’s redistricting plans retrogressive. as The pate Amicus Curiae drawing and the of one plans are statewide the court for necessarily has moved affects The ACLU district’s boundaries as amicus curiae Furthermore, participate to neighboring leave districts. position of defendants’ support case in majority- this from a removal of intervenors retrogressive. Plan is that provide minority district is sufficient to limitations opinion that the court is of standing challenge to intervenors by the Seventh on amicus filings outlined proposed district. Organization in National Circuit raises concerns that The State also Scheidler, 223 F.3d 615 (7th v.

Women btigation may in the intervenors’ interests Cir.2000), here. The ACLU applicable are diverge from the statements counsel. unique information presented no has testify appeared the intervenors to Two of in this can assist the court perspective that they prefer would depositions at their that matter, only to make additional and seeks Repubbcan majority-white, reside in a of the United arguments on behalf legal their This conflicts with counsel’s district. States, adequately represent a more than representations intervenors Accordingly, the court denies party. ed by a decrease overall harmed to file an motion for leave the ACLU’s strength caused reductions curiae brief. amicus How minority population districts. ever, interve- Standing Intervenors we are reluctant strike B. of Jones alleged solely on the basis of this nors in- challenged has The State of declara Intervenors’ sworn contradiction. reappor standing to contest the tervenors’ an caused clearly allege injury tions context of a Sec plans. tionment minority voting strength. diminution of Supreme challenge, Court tion may cast doubt on Deposition testimony plaintiff resides “[w]here stated injury, but it does the extent of district, ... racially gerrymandered standing. eliminate intervenors’ equal treatment plaintiff has been denied ra rebanee on legislature’s because of the C.King Motion for Reconsideration criteria, standing therefore has cial Intervene Motion to action.” United chabenge legislature’s 737, 744-45, 115 Hays, 515 U.S. States v. are two mo- before the court Pending (1995). 2431, 132 L.Ed.2d S.Ct. intervenor, Michael by putative filed tions at- African American King. King B. is an the four argues that two of State registered voter who resides torney and reside a bench- individual intervenors King pro- District par- Georgia Senate disputed by the mark se and first sought to inter- ceeding pro ties, Dis- removed from this but would be 19, 2001.5 On Deeem- vene on December redistricting plan. trict under the Senate motions). judgment summary responsibil- review for acutely of its The court is aware attorney, See, King is an Despite that Mr. e.g., the fact pro litigants. Fox v. Strick- to se ities King land, (D.C.Cir.1988) care to ensure (requiring this court has taken 837 F.2d 507 progress of this case. was aware of the pro plaintiffs of standard to inform se courts 21, 2001, 2002, King’s 15, ber motion was January denied On King filed a re- prejudice without because the United newed motion to January intervene. On yet 16, 2002, had position States identified its this court issued a scheduling respect plans with to the in redistricting directing parties order any to file and question and the court was unable to all responses King’s de- to motion by no later King’s 17, noon, termine whether interests would January than 2002 at and order- adequately represented by the existing ing any replies and all by be filed no 24(a). parties. 18, See Fed.R.Civ.P. Howev- later than January 2002 at noon. Also er, King granted permission 16, was 2002, to file an on January chambers contacted by amicus curiae parties brief no later than Janu- attempted this matter and to 14, ary 2002.6 manually Chambers faxed King contact order to inform them of King copy 21, of the court’s December the contents of the court’s order. Counsel 2001 order.7 parties for the they indicated that had not copy received a King’s renewed motion An initial scheduling pretrial and order telephone to intervene. King’s answering 20, this matter was issued on December system January was full on 16 and on the 2001. This order set forth a series following days. two In light of these cir- designed expedite deadlines to pro- trial cumstances, copy chambers faxed a of the ceedings. particular, the order re- scheduling 16, order King January to on quired defendants identify position to their 2002, and copy faxed a of King’s renewed respect plans to the redistricting plaintiffs motion to counsel for distribution 31, no later than December 2001. Cham- parties.8 to all manually bers a copy faxed of the initial scheduling pretrial order to King, the evening On of January light expedited of the pro- nature of the King contacted chambers and stated that ceedings, King’s failure to to subscribe he had received the facsimile copy of the automated faxing program of the Clerk’s January 2002 order. He noted that the office and representation his to chambers’ deadline for his passed submission had staff that he had not viewed filings just indicated that he had received thе this case. order. At no future King any date did file King upon receipt court contacted 8.Footnote 1 of Plaintiff's Memorandum of his motion to intervene and directed him to file his motion in accordance with the local Points and Opposition King's Authorities in to Intervene, Renewed Motion to January filed courts, governing three-judge rule which re- 17, 2002, reflects these events: quires parties pleadings qua- file all Although King’s Mr. Certificate of Service druplicate. See Local Civil Rule 9.1. indicates he mailed his "Renewed Motion date, King 6. To has not attempted filed or Intervene” mail to two of the file an amicus curiae brief. attorneys (only D.C. counsel and no Atlanta counsel) 14, 2002, January in this case on King copy Chambers faxed of the order undersigned represents counsel that as point King because at no did subscribe to the January close of business on faxing court’s program, automatic which al- 2002, none of the counsel in this case re- parties lows automatically receive facsimile King’s papers. ceived undersigned Mr. copies of court they orders at the time copy obtained a plaintiff, King’s docketed. of Mr. Counsel for Renewed federal defen- dants participate Motion via fax intervenors all from the law Court’s clerk faxing program Court’s automatic January and receive. immediately copies facsimile of all orders docketed in this copies faxed to other counsel in this case. case.

35 timeliness, solely disposi- it is not tion of mo- responses to his parties’ the reply to from is to be tive. Timeliness determined to intervene. tion circumstances.”). King “failed to all the scheduling initial our from As is evident timely fashion” interest a protect [his] order, matter was sched- this pretrial and with repeatedly failing to communicate by 1, February on to trial proceed to uled court, of the case and keep apprized the to statements, reports expert Pretrial 2002. filing requirements. local comply with to all filed Jan- testimony were direct and 367, King knew or Id. at 93 S.Ct. 2002, conference 18, pretrial the uary proceedings the known that should have 25, 2002. On January for scheduled was subject expedited to review. were 2002, an order 23, issued the court January ad- interest not been King’s legal has conference the rescheduling pretrial denial affected versely Court’s 25, January on p.m. to a.m. 12:00 10:00 (con- intervene. Id. at 368 motion to his copy of this order. 2002, King a and faxed future of movant to take sidering ability pretrial confer- appear at the King did interests). This court’s protect action to any way ence, did he communicate nor in no intervene King’s motion to denial of expedited concerning chambers with challenge ability to way forecloses his 30, January On for this matter. schedule plan. reapportionment Georgia’s senate to King’s motion 2002, denied this court expedit- consistently stressed We have This denial prejudice. without intervene In this matter. our review of ed nature of at appear “failure to King’s on based was to inter- renewed motion King’s denying conference, 25, pretrial January of the fact that vene, was court mindful notice, fail- his consistent had which he timely matter had to act a King’s failure with Chambers to communicate ure seriously disrupting for potential “the matter, in this parties counsel for the 369, at process.” Id. State’s electoral proceed- of these nature expedited 4 proceed- (discussing Section S.Ct. 01-2111, Order, Action No. Civil ings.” ings). 30, 2002. Jan. 7, 2002, a mo- King filed February On as a matter to intervene King sought court, this stay proceedings tion to 24(a). to Fed.R.Civ.P. pursuant right to of his motion reconsideration motion for However, for any application intervention appeal to the intervene, and a notice of National Ass’n timely. See must for stay for a The motion Supreme Court. New People v. Colored Advancement of 4, hearing,” “January to a referred York, 93 S.Ct. 413 U.S. stay pending moved to King which (“NAACP”). (1973) L.Ed.2d re- motion for of his court’s consideration NAACP, Supreme Court discussed hearing No appeal. his consideration timeli- considering the for legal standard January this matter was scheduled action intervene an a motion ness in this trial commenced court 2002. The pursu- brought declaratory judgment possible it is 2002 and February case on 4(a) Rights Voting of the to Section ant stay request King intended 365-66, 93 S.Ct. 413 U.S. Act. Nevertheless, filed on was the motion trial. three-judge court held that Court four- day of the the last February deter- discretion in its exercised properly day trial. the circumstances from all of mining that, re upon argues Plaintiff Id. untimely. was to intervene motion lost this court appeal, the notice ceipt of which the suit point (“Although King’s motions. to consider jurisdiction in the determina- one factor progressed agree. Supreme We The pro- Court has ing are unduly prejudicial. Epstein vided clear direction proceed as to how to relies on the table in question only for the when a defendant simultaneously files a limited conclusion that a probit statewide notice appeal appellate with the court analysis proper; was neither he nor the and a motion for reconsideration with the suggests State table reliable district filing court. “The of a notice of evidence of white crossover voting in the appeal jurisdictional is an event of signifi- Senate Districts. jurisdiction cance—it [ap- confers on the urge Intervenors this court Ep- to strike pellate and court] divests the district court stein’s testimony on the basis that it is not of its aspects control over those of the case competent expert testimony. They argue appeal.” involved in the Griggs v. Provi- Epstein’s probit analysis rep- does not Co., dent Consumer Discount 459 U.S. resent reliable relevant evidence. How- 103 S.Ct. (1982); L.Ed.2d ever, Epstein testified probit analysis DeFries, accord United States v. 129 F.3d is a standard statistical methodology. The (D.C.Cir.1997). itWhile is clear to court Epstein’s finds that report is reliable King’s court that renewed motion to and relevant evidence. intervene and stay motion to are as un- timely previous motion, as his King’s no- Findings II. of Fact appeal tice of juris- divested this court of glance, At first the evidentiary record in diction to King’s consider motions. this appears matter Yet, extensive. con- D. Motion Epstein’s to Exclude Dr. Tes- sidering that the State pres- has chosen to

timony ent three reapportionment statewide plans court, to the record fact is rather

The United States and intervenors slim. The of Georgia, State the United seek to exclude the testimony of the States and intervenors have all contributed expert, State’s Dr. Epstein, David regard to the evidentiary record ing before court. his analysis white crossover voting The State introduced statistical data on Districts, benchmark Senate see the existing proposed districts, U.S.Ex. includ- his conclusion that it was ing political proper pеrformance, population total assess trends on a state wide, voting age populations, as well regional rather than a basis. Al break-downs of though that data testimony race. was provided at the State relied eleventh on the hour States, testimony expert to the two United witnesses, legislators, United State States was United able to States cross-examine Representative Epstein respect with John to his Lewis analysis. Georgia, and the Epstein cross-examination of director of the effectively redistricting office highlighted problems Georgia, Epstein’s Meggers. Linda response, con clusion that there was no United statistically presented States sig the court with nificant greater variation in degree amount of white of and more detailed Furthermore, evidence, crossover. including court data, offered voter registration the United precinct-level States an opportunity information, to re maps data and open cross-examination in permit demonstrating exactly order how district lines the United expert States’ to assist counsel would be redrawn plans, in the cross-examination. The United testimony numerous social leaders States declined this opportunity. The and local elected officials from the contest- court finds that Ep introduction of ed districts. The United pro- States also stein’s calculations of white crossover vot- vided the only expert report that consid- map is the remedial This court-drawn polarized racially of prevalence the ered this court’s consider- plan9 for evidence benchmark States’ But the United voting. complaint for declarato- plaintiffs of ation scope focusing limited extremely was — of Findings Proposed judgment. Pl.’s ry contested districts three only on (“PPFF”^ 53. of Law was Fact & Conclusions That evidence plan. Senate State assess the court to permit to designed and Sen- House State The current State plans. three of of each impact overall result of a place as a put were plans ate all challenged intervenors Finally, while by the adopted plan, which was mediated evidence present little they plans, three Assembly in John- 1997. Georgia General and plans alternative other than 1561-67; Miller, F.Supp. at 929 v. son ex- the State’s critiquing expert report an ¶ to a settlement 68. Pursuant PPFF report. pert lawsuit, both of in a 5 section agreement to plans were submitted the current Elections Reapportionment A. pre- and were of Justice Department Background Georgia: 29, File DOJ April 1997. See cleared on ple- Assembly Georgia General of (granted as reconsideration No. 95-3656 authority under Constitution nary ¶ submission); PPFF 68. 1995 original enact, Georgia of of the State laws of the adoption Since or veto of approval subject to the minor 1997, amended it has been Governor, reapportion legislation and those amend- three times changes Representa- and House of State Senate De- U.S. precleared ments were Georgia’s designated tives, as of as well (DOJ 23,1998 April on partment of Justice House in the U.S. number seats 98-0912), Sep- File No. 98-98-0759 Ill, Const., Art. Ga. Representatives. No.1999-0989) (DOJ 20, File tember 21-2-4, 21-2-3, ¶ §§ II; II,§ O.C.G.A. (DOJ 28, 2000 File No.2000- August fact, Georgia’s State 28-2-1; 28-2-2. ¶ 2682). of the adoption Id. 69. Since the General mandates Constitution 1997, amended it has been plan in House and the reapportion the Senate Assembly precleared was change Id. That once. necessary af- as Representatives House April on of Justice Department the U.S. census. States decennial each United ter (DOJ and 98- File No. 98-0759 Ill, II, Const., § 2. The State Art. Georgia amended, 0912). plans, Id. as the dis- provides that further Constitution for the plans the benchmark constitute contiguous ter- composed of tricts shall be redistricting House and State State ritory. Id. consider- to the court for submitted plans Congression- United States The current ation. court-drawn the result al districts are census, the national Following place after put which was map, remedial redistricting enacted legislature Georgia decision impasse and court legislative to take are intended plans. plans These un- were Congressional districts that two general by the time next effect in the effort constitutionally based race Tuesday, No- for day, scheduled per- election minority population increase their Georgia time at which Johnson, vember 521 U.S. v. centages. Abrams to the United (1997). candidates will elect 138 L.Ed.2d 117 S.Ct. plan. able existing district- plan is The benchmark effect, legally enforce- or the last ing plan in Congress and the Georgia (“VAP”) States General total voting age population ¶ Assembly. 21-2-9; ¶¶ O.C.G.A. PPFF in the state. Id. 26. The total VAP ¶ upcoming In the cycle election non-Hispanic individuals identifying Congress and the General Assem- themselves was 24.46% of the partisan bly, candidates for offices will ¶ total VAP. Id. 26. qualify for either the Party Democratic or 2000 The decennial census shows that Republican Party nomination between 9:00 population the total of the State of Georgia n .m. on June 19, 2002 p.m. and 12:00 on 1,708,237 has increased by individuals 21, § 21-2-153(c)(l); June 2002. O.C.G.A. 1990, 8,186,453. since and is now PPFF ¶ primary The

PPFF 13. for nomination ¶ 6,017,219 27. There are people Geor- partisan office in Georgia will next be ¶ gia who are of voting age. Id. 31. August held § on 2002. O.C.G.A. 21- ¶ 150(b)(1);PPFF 14. Any The 2000 run-off elec- census allowed individuals for 2— tion necessary August the after first time to primary identify themselves as election will on Tuesday, be held Septem- more than one race. As a result of this § ber 2002. O.C.G.A. 21-2-501(g); change, parties dispute proper cal- ¶ PPFF 15. of culation the African American popula- tion Georgia. In special the 2001 re- In Georgia, a seeking candidate nomina- session, districting the State of Georgia tion to a state or federal office in a regular defined “black” as including non-Hispanic partisan primary must receive a majority Hispanic black persons of a single of the votes cast in the primary inor race, and “black persons combo” as all primary run-off election. § 21- O.C.G.A. ¶ who identified 2-501(a); themselves as black in PPFF 18. A candidate is combination with any other elected to office racial regular or general eth- elec- nic upon receipt category tion of a plurality 45% 2000 census vote. form. §§ O.C.G.A. U.S.Ex. 501(g),21-2-2(22); 18:13-24. Consequently, Id. 21—2— ¶ purposes 19. In the matter, event no candidate this receives Georgia has plurality, such a runoff its population election counted black is then as including 21 days held later. Id. all black Hispanic multi-racial and non- Hispanic responses. contrast, the De- B. Demographics partment Justice, in accordance with a The 1990 Guidance issued by census showed that Department the total population January, of the State of Georgia was counted as black those n ,478,216 ¶ persons. non-Hispanic PPFF 21. The 1990 identify individuals who as 1,746,565 census also showed that black only, or persons white, as black and but not Georgia, 26.96%, or them- individuals who identified identifiеd as black and an- ¶ as 1,737,165 selves black.10 Id. 22. per- other race. See Guidance Con- sons, or 26.82% identified cerning themselves as Redistricting and Retrogression non-Hispanic and only. Id. The 1990 Under Section 5 of Act, Voting Rights census reflected that black voting age pop- § 1973c, U.S.C. 66 Fed.Reg. (“BVAP”) ulation 1,168,142, (Jan. was 18, 2001).11 or 24.58% appears It 1990 and 2000 opinion, census 10. this section will use "black” to reports provided by parties identify that the responses cen- individuals’ to the census. respondents sus asked identify themselves court Georgia's will refer to method of Thus, as "black.” while we refer to "African counting (Ga.)”, BVAP as "BVAP and to the American” throughout individuals and voters (U.S.)”. United States’ as "BVAP Id. of the VAP. represent 0.05% category Population Total ¶ 36. 2,349,542 showed 2000 census themselves Georgia identified residents may identify their voters Registered of the 28.7% only, representing as black However, to vote. they register when race ¶27. The PPFF population. state total census, per- are not unlike the identifying population non-Hispanic total one race. identify as more than mitted to 28.5% only represents as black themselves general time of November At the Id. 0.22% population. total state of the 3,177,061 regis- people election there were themselves identified population total 698,305 these, Georgia; tered to vote in only, 0.21% and white as black 21.97%, as themselves identified people, themselves identified population total ¶ In November Id. black. only. white and black non-Hispanic and to vote 3,003,527 registered were people Id. as of whom identified Georgia, 21.92% identify- Georgia of population ¶ The total In November 38. Id. black. racial other and one black ing as vote iden- persons registered to 24.38% of population; the total 0.47% of category ¶ was 39, and, black, id. as themselves tified as identified population 0.20% of the and regis- persons of 24.73% November cate- racial non-white one other and ¶ black 40. as Id. black. tered to vote identified ¶ of the additional 0.07% 29. An Id. gory. general 2000 November the time of the At identified as population total to 3,856,676 registered election, persons Id. category. racial other than one more 980,587, or of which Georgia, vote ¶ 30. black. Id. 25.42%, as themselves identified Voting Population ¶ Age Total 41. (“VAP”) population voting age

The total Reapportionment Process The 2001 as black C. identifying themselves Georgia of of the 1,602,985people, or 26.64% only was of the Gen- session During the 2001-02 ¶ popu- voting age 32. The Id. total VAP. in the 180 seats there were Assembly, eral identify- non-Hispanic individuals of lation and 56 Representatives House of Georgia only is 26.52% as black themselves ing During the Georgia Senate. seats Findings of Proposed total VAP. U.S. Georgia General 2001-02 session (“USPFF”) of Law Fact & Conclusions Representatives Assembly, 34 ¶ identifying as black 124.12 The VAP African 11 of the 56 Senators were VAP; the of the 0.08% represents white the African All of American. VAP is and white non-Hispanic black total are Demo- and Senators Representatives ¶ 34. Those PPFF 0.07% of total. 17:13-16; dep. Eric crats. Johnson and one as black identifying themselves at 17:18-24. dep. Lynn Westmoreland 0.28% of constitute category racial other 22, 2000, of the the results March ¶ On identifying VAP, those id. while for became decennial census and one non-Hispanic black themselves Bu- the Census generally available 0.25% of comprise category racial other Assembly to reau, the General prompting themselves identifying Id. VAP. Those ¶ formed The Senate PPFF racial act. other than one and more as black 1,591,421, or fact, Georgia is Hispanic VAP in PPFF finding this 12. Plaintiff admitted 6; ¶ id. n. but see n. however, PPFF 33; findings ¶ plaintiffs 26.44%. own percentages proper (suggesting that fact, believes the that it states footnote it 27.83%). responses for non- of census proper number reap- committee address the issues of House and Reapportionment Senate Com- portionment.13 During the 2001-02 ses- joint mittees held hearings as follows: sion, chairman, vice-chairman and sec- April Watkinsville; April Atlanta; 17— 18— retary of the Reapportionment April Augusta; April Perry; May 25— 30— Committee were Tim Golden, Senators Brunswick; Valdosta; May May 10— 15— Robert Brown and Hugh respective- Gillis ¶ Dahlonega. Id. 71.14 23— ¶ ly. Id. at During the 2001-02 session prior Also special session, of the Assembly, General that Committee the House and Reapportionment members, had six of whom were African adopted Committees guidelines ¶ providing American. Id. at 7. Those six were public for access to committee hearings Vice-Chairman Brown Robert of the 26th district, and meetings, public access Senator 15th, redistricting Harbison Ed data 36th, Senator David general Scott of the materials and guidelines Senator 10th, Nadine Thomas of Regi- presentation Senator and introduction of ¶ na Thomas of the 2nd and Majority plans Lead- to the committees. Id. 72. er Charles Walker the 22nd. Id. The Senate Reapportionment Commit- Representatives House of also tee met formally April June July formed a committee address the issues 12, 1, 6, August 9, August August August of reapportionment. chairman, The vice- 27, 28, August 4, September September 7 chairman and secretary of the Leg- House ¶ September 13, and 2001. Id. 73. Tran- islative Congressional and Reapportion- scripts or records of those proceedings ment Committee Representatives were provided were Department the of Jus- Smith, Tommy Jay Shaw and Carl Von part tice as a of litigation. this Id. The ¶ Epps, respectively. Id. 8. During the Legislative House Congressional and Re- 2001-02 session of the General Assembly, apportionment 11, April Committee met on the House Reapportionment Committee 29, 10, July 20, June July 24, July July 26, members, has 29 six of whom were African 31, July 13, August 14, August 16, August ¶ American. Id. 9. Those six were Carl 22, August 28, August 29, August Septem- Epps district, Von of the 131st the Secre- September ber and 10. Id. Transcripts tary of Reapportionment the Committee or records of those proceedings pro- were and Chairman of Legislative the Black vided Department to the of Justice. Id. Caucus, 124th, David of Lucas the Lester 148th, Jackson of the Arnold Ragas 21, 2001, On June Roy Governor Barnes 64th, 52nd, Kasim Reed the of and LaNett proclamation issued a calling the General Stanley-Turner of the 50th. Id. Assembly special into purposes session for Prior special to the 2001 sessions to of reapportioning the State Senate and consider reapportionment issues, ¶ the House of Representatives.15 Id. 74. legislative 13. The official website for the 2001 15.Regular Georgia sessions of the General redistricting process by the Georgia Assembly General on the Monday commence second Assembly January is of year may each http://www.legis.state.ga.us/Legis and continue in period session ¶ for a longer of no 02/reapp/index.htm. than PPFF 72. /2001— Const., legislative Ill, IV, days. § Ga. Art. I;¶ ¶ PPFF 10. Georgia The Governor of object Intervenors that these were 14. may convene the Assembly special General "hearings” but do otherwise not contest the by proclamation session and the laws dates and locations. Intervenors’ Pre-trial may during special enacted such a Proposed Findings Fact & of Conclusions of session are those relate purposes to the ¶Law 116. proclamation stated in the any or in other 2001, by the and August Senate on gia of the General session special first This Au- on Representatives of Georgia House on 1 and ended August on Assembly began House redistrict- 2001. State gust Id. 2001. August Georgia House by passed was ing plan Geor- worked with Meggers Linda and August on Representatives since Redistricting Office Legislative gia’s September Georgia Senate by the ¶ with full-time worked 76. She Id. redistricting Congressional 2001. The has served and office since houses passed was both intimately Id. She since Director September Assembly on Georgia General changing demographics, with familiar Republicans either 28, 2001. No geography political demographics, any of these for voted House 10-16. 22 at Pl.Ex. entire state. (Eric Johnson plans. reapportionment testimony, direct provided Meggers 27). African American All of dep. the United deposed was twice she Assem- General legislators States. ¶ Democrats, PPFF 5. With bly an overview gave Meggers Ms. Repre- American one African exception of testifying process, reapportionment Sena- African American and one sentative differences significant there were against the State House voted who tor compared process redistricting this plans, African redistricting State technology Id. data years. past redis- voted for the legislators sophisticated in 2001 allowed available tricting plans. performance political analysis of goal that the districts: Meggers testified prospective Ms. leadership the Democratic we the exactly is what geography Political two-fold: House was here, dis- and new Senate talking about districts, dis- House tricts, congressional minority dis- the number To maintain we So tricts; political geography. that’s had, at the but presently we tricts that district, House could draw the maintain increase same time the and have geography, piece of House they seats Democratic number of it immediately analyze data, and census They the Senate. House and had existed had If that district politically. maintain just they couldn’t knew that how it would this is 1996 or had, needed to they actually they what elections, particular these voted in they if were majorities those strengthen had the data. we where decade. majority over the maintain assess available to was 9. Data Id. leadership, you Democratic say IWhen Demo- for to vote tended districts whether Cau- Black to understand need Id. past elections. Republicans crats Black Caucus members cus that, in contrast testified Meggers Ms. in that. very involved were leadership data elections, performance political past leader- of the part much a very They are the 2000 redistrict- extensively in So, used they was this. we talk about ship when at 17. Id. process. districts, ing but those maintain wanted to I term waste, guess the is the—I plan be- redistricting The State often, their votes. waste heard Geor- approved was this court fore in accor- II, days extended Const., V, unless legislative § Art. thereto. Ga. amendment KVII(a); Ga. Constitution. Georgia ¶ dance with Special sessions PPFF ¶ ¶ V, II, VII(c); Const., PPFF § Art. period limited to Assembly are General PLEx. at 20-21. One of the Representatives reasons House of pro- and Senate given by African American senators for duced Congressional plan redistricting aligning their interests with those of the submitted to this court preclearance. that, Party Democratic was (Brown should test.). the Pl.Ex. 20 at 28 The Confer- Democratic Party in major- cease to be ence members, Committee had six three ity in the State House Senate, and State from the Senate and three from the House existing all American African chairs of Representatives. six, Of these two Sen- committees would lost. dep. C. Walker ators and Representative one are African Brown, at 94. Senator Robert an African American. Id. at 27. from Senate District was Vice Chairman of the Senate Reapportion- 1. The Benchmark Plan overall, ment Committee and was According results, to the 2000 census chairperson of the subcommittee that did percentages population (“BPOP”), of black the Senate Plan itself. Pl.Ex. 20 at 23. black age (“BVAP”), population Brown, According to Senator there are 11 (“BREG”) registration black for each of

African Senate, Americans in the state Georgia’s existing Congressional districts 7 to 8 of that currently number could chair under the benchmark are as follows: committees. Id. at 18-24. The majority BVAPCGa.)16 %BPOP % bBREG leader of the Senate is an African Ameri- can, and the chairman of the rules commit- District 31.65 28.97 26.23 District 40.85 37.38 35.68 tee, Brown, Senator is also African Ameri- District 3 31.27 28.62 26.69 can. Id. District 4 50.60 46.24 49.13 District 5 63.57 58.85 60.31 District 6 11.39 10.80 9.23 D. Congressional United States Redis- District 7 18.66 16.88 15.99 tricting District 8 32.66 30.28 27.65 District 9 3.40 3.12 2.53 After census, each decennial United 39.00 36.12 33.72 District 11 15.01 13.64 12.1 States Representatives House of reap- 8D, portioned Pl.Exs. population 8E. reflect current changes Congres- plan, sional the states. census, After there the 1990 are two districts over State of 50% total population, was assigned 11 seats pursuant Districts, Fourth and Fifth reapportionment. but PPFF one ¶ 101; district, 8A, District, the Fifth Pl.Exs. with 8C. The State over of Geor- 50% BVAP gia then had and black the responsibility registration. voter redistrict Id. However, District, to reflect the those 11 Fourth seats. As has discussed over above, 45% BVAP and redistricting voter subject registration. was Id. litigation that resulted in a court-ordered

redistricting plan. Johnson, Abrams v. Notwithstanding the fact that the Fifth 82-85, 521 U.S. at 117 S.Ct. 1925. The Congressional District is the only existing plan court-ordered that resulted from that majority African American district in litigation is the plan benchmark before this terms of voting age population and regis- ¶ court. PPFF 102. voters, tered the Georgia State of is cur- Following census, the Georgia 2000 rently was represented three by African Amer- apportioned 13 seats the United States Congresspersons: ican Sanford Bishop ¶ House of Representatives. (Second Id. 109. A District), Cynthia McKinney Conference (Fourth District) Committee of the Georgia (Fifth and John Lewis 16. percentages These reflect BVAP, as calcu- Georgia. lated the State of by 8,186,- 2; Tr., 2/4/02, upon population a total statewide District). p.m. 21 at Pl.Ex. the of 13 seats people, assignment at 59-61. Representa- House of in the United State currently Lewis John Congressman Georgia, the the ideal tives for State Congressional Fifth Georgia’s represents pur- for Congressional size of a district States House in the United District principle of one poses of adherence to 21 at 2:5-6. Con- Pl.Ex. Representatives. 629,727 people. Id. person, one vote is this one of Lewis been gressman ¶110. for rights civil advocates country’s leading actions, along and his past years, population statistics re- upon Based Jr., King, Dr. Martin Luther those of census, all ported Georgia’s in the 2000 passage achieving were instrumental congressiоnal districts have existing 1965. Id. at Rights Act of Voting than the ideal district populations larger 629,727, appor- are thus out of size of McKinney, current African

Cynthia percentages tionment. The which for the Fourth Congresswoman American current districts exceed ideal district District, elected in was first Congressional follows: size are as Dis- Congressional Eleventh District 1: 9.92% District 2: 3: 3.28% district was the BVAP of that trict when District District 4: 24.13% Tr., 2/4/02, upon p.m. at 61. Based 60%. 18.33% by the three- map remedial drawn 2.82% District 5: District 6: 7: Johnson, court, Abrams v. judge see 49.51% District 19.44% 83-85, McKin- 117 S.Ct. Ms. U.S. at 5.25% District 8: in the Fourth Con- ney ran for election 29.32% District 9: District, and was successful gressional 10: 5.16% District 1996, 1998, and 2000 when the BVAP 11: 32.82% 33%, 45%, respec- was 39% and the district ¶ Id. 111. ¶ 25, App.; PPFF tively. Pl.Ex. Proposed Plan 2. The African Bishop, the current

Sanford Congressman for the Second session, special During its second District, was first elected Congressional Assembly enacted Senate Georgia General was the total BVAP a district which 1EX2, reapportion- set forth the Bill which 2/4/02, Tr., Follow- p.m. at 59-60. 52%. congres- 13 new Georgia’s ment district lines ing redrawing of those 9A; Pl.Ex. PPFF districts. See sional 1996, Congressman court in ¶ the federal adopted by Bill 1EX2 was 112. Senate Dis- Bishop won reelection in the Second September Senate on State total was between trict when the BVAP ¶ Id. 114. No by a vote of 30 to 23. 25, App. In the Pl.Ex. 35% and 37%. Black Legislative the member of career, political Congressman of his course The plan. Id. against Caucus voted on Bishop Congress has won reelection to House adopted by was bill majori- in rural separate occasions three day, by on the same Representatives ¶ PPFF ty-white district. ¶ No member of 99 to 59. Id. 113. vote of voted Legislative Black Caucus the House reapportionment man- By virtue of the plan. Id. Governor results, against census State by dated the 2000 law Bill 1EX2 into Georgia.signed Senate in the United representation Georgia’s 1, 2001, Id. Act No. 2EX11. in- on October Representatives was House of States ¶ ¶ 109. Based to 13 seats. Id. creased

44 proposed Congressional plans.

Under the re- Br. of See Amicus Curiae/Defen- dant-Intervenors, 14, 2002, population distrieting plan, total at 26. Jan. (“TPOP”) voting age population and proposed plan would also create (“TVAP”) of each district is as follows: significant additional districts with African TPOP TVAP (1) populations: proposed lation tration Id. 116. sional Under the District District 1 District 2 District District District District District 5 District District District 7 District 11 District 12 ¶ plan, (“BPOP”), BVAP, (“BREG”) ^{[4] [13] [9] [8] [10] [3] [6]} the percentages proposed for each of the 629,732 629,748 629.762 629,706 629,702 629,698 629,725 629,700 629,727 629,735 629.761 629,735 629,690 Georgia Congres- of black black proposed 463,958 444,493 457,971 455,805 492,438 465,459 450,756 470,201 467,232 456,300 464,632 455,164 472,785 regis- popu- their registrations. PLExs. 40% and and 13th rican American candidates have announced black voter tration of Thirteenth Congressional District has Districts have black dep. at 107:11-18. the Third and (3) Second excess of 40% and a black voter intentions to Congressional significant nearly Congressional registration 40%; Twelfth run BVAP District has a BVAP populations 9C, for the new 12th of over (2) districts. Tate 9D. Congressional Several black voter 40%; proposed of over regis- Af- presented The State an analysis of the Congressional districts are as follows: District District District 7 District District 12 District 11 District District 5 District District 4 District District 1 ^[2] %BPOP 29.10 n 41.97 43.19 23.21 40.32 54.69 14.07 45.22 56.92 12.95 7.36 3.65 7.43 %BVAP(Ga.)17 38.22 26.36 39.00 50.02 37.55 21.04 41.46 52.04 12.99 12.07 3.36 6.81 6.87 %BREG 26.14 41.57 39.10 51.16 34.97 11.16 39.99 10.37 53.36 18.62 2.89 6.02 6.27 American candidates voters of the PLExs. District would between an African American candidate 70.83% 68.97% a Democratic primary, and 2000. This and a white candidate held between 1998 statewide election returns in four 9D, - 75.65% the 10B. proposed analysis predicted Fifth Congressional supported general an estimated elections. elections that the African 9C, PLExs. 9D. The State did expert not introduce testi- Under the Congressional re- mony interpreting significance of these districting plan, there are still percentages. Furthermore, two Con- court districts, gressional Fifth, the Fourth and expert heard no testimony regarding the majority populations, but the existence of racially polarized voting pat- number of Congressional districts with terns in any of the benchmark or (Ga.) over 50% Districts, BVAP and black voter reg- Congressional impact or on the one, Fifth, istration has increased patterns of such ability minority two, the Fourth and Fifth. Id. Howev- candidates to win election. Intervenor- er, as intervenors emphasize, according to defendant Patrick L. Jones testified that BVAP, United States’ calculations of “difficult, he “believes” that it is if not there is one district with a majority impossible” for candidates of BVAP—the Fifth District —in both the choice to be elected in districts of less than benchmark BVAP, Congressional 55% and that it will be difficult to table, purposes 17. For this Georgia. we use BVAPas calculated the State of person, one vote district for one in the Fifth House of choice candidate elect a 45,480 people. PLEx. 12C purposes is Int.Ex. 27. District. Congressional ¶ 8; PPFF 142. alternative have submitted Intervenors population According to the 2000 census increase of which would some plans, statistics, there are 40 districts See majority-minority districts. BVAPs plan in the total non- benchmark which However, alter- the three Int.Exs. 20-22. population 50%.18 Hispanic black is over cre- by intervenors plans submitted native ¶ addition, 37 dis Id. there are districts Congressional two ate at most plan tricts in the benchmark which the the alter- majorities. None of BVAP ¶ This is over 50%.19 Id. 144. total BVAP incumbents place Republican plans native true whether BVAP is calculated accord in the same district. Attorney Guidance or ing to General’s Redistricting E. House State methodology. Finally, under by Georgia’s there are 38 districts plan, the benchmark mandates Georgia Constitution registration in which the total voter con- Representatives Georgia House 50%.20Pl.Ex. 11E. is over appor- members not fewer than 180 sist of of the State of among districts tioned districts have tradition- Georgia’s House ¶II, 1(b); Const., Ill, § Art. Georgia. plus Ga. a deviation of ally been drawn with ¶ 28-2-1; PPFF 135. Mem- § O.C.G.A. the ideal district percent minus five ¶ Representa- Georgia House of bers of the to the 2000 According PPFF size. terms and two-year results, for existing tives are elected only two of the census of the convening time of the popula- until the serve with a total black House districts Const., Art. Assembly. BVAP, Ga. tion, registra- next General or black voter total ¶ ¶ II, V(a); Ill, § PPFF 136. Members 50%, traditional fall within that tion over 11D, Representatives Id.; House of Georgia of the PLExs. requirement. deviation time as the Gover- majority are elected at the same All five of the 37 BVAP 11E. but - ¶ Const., V, I, II; § PPFF Art. nor. Ga. were between -7.23% districts ¶ 137. district in deviation from the ideal 31.92%

size, they significantly were indicating Plan 1. The Benchmark Id. underpopulated. plan Georgia for the The benchmark Proposed Plan 2. The contains 180 sin- Representatives

House ¶ session, Georgia 140; special PI. In its first PPFF gle-member districts. redistricting Assembly passed a HA, upon popula- a total General Exs. 11C. Based that was not the State House 8,186,453 existence people and the tion of dep. at 89: by the James signed House of Governor. of 180 members (Westmoreland 4-12; Int.Ex. 31 at 3-4 the ideal size of State Representatives, 148, 127, 133, 134, 136, 140, 120, 124, 48, 121, House Districts 18. These are benchmark 58, 64, 65, 11D; 52, 53, 54, 55, 56, 57, 161, 49, 50, 51, 149, 151, Pl.Ex. PPFF and 162. 116, 117, 71, 72, 73, 75, 93, 96, 66, 68, 69, 70, ¶ 144. 127, 131, 133, 134, 136, 120, 121, 124, 118, 11D; 148, 149, 151, 140, Pl.Ex. 161 and 162. 48, Districts benchmark House 20.These ¶ PPFF 143. 55, 56, 57, 58, 64, 52, 53, 54, 49, 50, 51, 116, 117, 72, 73, 75, 93, 96, 66, 68, 69, 70, 71, benchmark House Districts 19. These are 140, 148, 127, 133, 134, 136, 118, 120, 124, 58, 64, 65, 55, 56, 57, 49, 50, 51, 52, 158, 161, and 162. Pl.Ex. HE. 66. 68. 69. 70. 71. 72. 73. 93. 116. 117. 46

Decl.). area, leg- African passage After of the first were Fort, plan, Vincent who is against plan. House Senator who voted Id. islators American, meeting called for a African H.B. 14EX2 into law signed The Governor Legislative Black Georgia Caucus 1, 2001, Act on October No. 2EX23. Id. (“GLBC”). 24, ¶ In an 2001 letter August 149. Representative Carl Von addressed proposed plan contains 180 This House Caucus, Epps, of the Senator Chairman members allocated to 147 districts. PLEx. Fort stated: ¶ 12C; PPFF 150.124 districts contain one are concerned that the GLBC has We member, members, 15 districts contain two in redistricting not been involved members, six districts contain three resulted, all. process almost at This has two districts contain four members. Id. plan among things, legislative other proposed 42 plan House contains passing majority-minori- that has diluted population districts which the total black ty districts both the House and the is over PLEx. 50%. 12C.21The Senate. plan House contains 39 seats districts sig- Int.Ex. 17.- This letter contained the BVAP, pursuant which the total to the Id.; natures of six members of the GLBC. data, interpretation State’s of the census U.S.Ex. 51:23-52:16. Senator Fort over 50%.22 PLEx. 12C. When BVAP is spoke either called or each member of pursuant to Attorney calculated Gener- signature appears upon

the GLBC whose Guidance, plan al’s the redistricting con- letter, and each signing consented to tains 38 House seats in which the BVAP is 55:22-56:10; the letter. Id. at USPFF at ¶ U.S.Resp. over 50%. to PPFF 152. The ¶ 64. House contains 39 districts session, During special its second registration which black voter is over Georgia Assembly General enacted House PLEx. 50%.23 12D. 14EX2, Bill provided reap- which for the Comparing proposed plan to the portionment Georgia Rep- House of plan, benchmark there are two additional resentatives. See Pl.Ex. 12A (identifying ¶ “HSEPLN2”); populations districts with black of over plan as PPFF 147. 50%, and one additional adopted House Bill 14EX2 district with black was 12C, registration voter Representatives House of over 50%. Pl.Exs. on Au- 29, 2001, gust by a vote of 12D. Either one or two additional seats are 100 to 72. Id. ¶ passed BVAPs, 148. The created in Sep- majority the bill on districts with 6, 2001, by depending tember a vote of 29 to 22. on whether Id. the United States Representative Dorothy Georgia’s Pelote and Sena- or calculating method of BVAP Thomas, ¶ Regina Id.; tor both of the U.S.Resp. Savannah is used. to PPFF (two-member 62, 81, 83, 97, 98, 100, 105, district), 21. These are House Districts 43 member district), 44, (four-member district), 107, 111, 113, 114, 48 (two-member 124 dis- 49, 50, 51, 55, 57, 58, (three-member 59 dis- trict), 135 and 136. Pl.Ex. 12C. trict), (three-member district), (three- 60 61 62, 81, 83, 97, district), 98, 100, 103, member (two-member 23.These House Districts 43 105, 107, 111, 113, 114, (two-member 124 44, 45, district), (four-member district), 48 district), 125, 128, 135, and 136. Pl.Ex. 12C. 49, 50, 51, 55, 57, 58, (three-member 59 dis- trict), (three-member district), (three- 61 (two-member 22. These are House Districts 43 62, 81, 83, 97, 98, 100, 105, district), member district), 44, 45, 47, (four-member district), 107, 111, 113, 114, (two-member dis- 49, 50, 57, 58, (three-member dis- trict), 135 and Pl.Ex. 12D. trict), (three-member district), (three- 9.17%, Challenged short of the ideal district size. Pl. The Districts 11D, registration Exs. 13B. Black voter challenge drawing Intervenors levels, compared regis- to overall voter 97, 100, 113, districts, 51, 95, House seven tration, have declined in benchmark House 125, 124, as the creation of as well past District 56 over the three election Renewed multi-member districts. See 1996, cycles, from 58.91% in to 58.74% in Intervene, 4, 2002. The al- Jan. Mot. finally 55.86% 2000. Pl. plans reapportionment House ternative HE, Exs. 13B. intervenors, and drawn submitted Lynn Republican House leader Westmore- Voters within benchmark House District land, incumbents place certain Democratic Party 56 tend to vote for Democratic can- run one against in the same districts to didates, by an as demonstrated overall See, e.g., Int.Ex. 31 at 36:19- 86.92%, another. performance Democratic score of 38:17-39:11, 44:4-25, 46:12-16; 37:18, performance as well as the Democratic 53:2-54:18, 58:15-19, 48:11-16; 50:1-51:16, years numbers for the individual election pro- (82.14%) None of Westmoreland’s (80.25%), 59:5-60:2. of 1996 and 2000 plans (81.50%).24 alternative drew districts posed HE, PLExs. 13B. Republican incumbents were drawn which addition, In 60.9% the voters within 43:3-8; district. Id. at within the same the benchmark House District 56 voted for 54:19-22; 48:17-21; 51:17-23; 62:17-22. Thurmond, American, Michael an African Tyra objects to the Intervenor Roielle opponent over his white the 1998 Demo- majority-minority House dis- loss of one Party primary cratic runoff election for the single district where resulting trict and open State Labor Commissioner seat. “pitted against two members 88.43% of voters benchmark House Dis- However, when each other.” Int.Ex. general for Thurmond in his trict 56 voted whole, proposed House reviewed as victory Republican election over a white opportunities for plan creates four new HE, Additionally, opponent. PLExs. 13B. Americans to elect candidates of African House Dis- voters within benchmark open choice in seats Dis- their House support trict 56 demonstrated electoral for opportunity tricts and a new candi- other African American Democratic 12B, 12D. in House District 60. PLExs. office, voting running dates for statewide Attorney Baker for 1998 for Thurbert Proposed District 51 a. House and, in General at a rate 82.21% for the Burgess, David a candidate Proposed single- House District 51 is a Commission, at a rate of Public Service Ful- wholly district located within member HE, 13B. 12B, 79.23%.PLExs. County, 13A. Georgia. ton Pl.Exs. territory proposed district embraces included in the precincts in benchmark House formerly included supported House District 51 have Demo- wholly is also within which past elec- According cratic candidates. 11B. County. Fulton PLEx. results, precincts comprise tion results, 51 have an overall House District light of the 2000 census of 86.38%. performance is Democratic score

ideal size for one of the 180 House seats results, Thus, 45,480 Using PLEx. 12D. 1996 election persons. the benchmark 4,169 perfor- persons, plaintiff projected or Democratic ‍​‌​‌‌‌​‌​​​​‌‌​​‌‌​​‌‌​‌​​‌‌‌‌​‌‌‌‌‌​‌‌‌‌‌​‌‌‌‌​‍District 56 is House supported given precincts district performance in- 24. The Democratic numbers candidates. degree voters in the Democratic dicate the to which the 80.38%,using Additionally, sup- manee numbers of 1998 elec- 81.55% of these voters Baker, 81.82%, ported Thurbert and 78.67% voted results, using tion 2000 elec- Burgess. for David Id. results, 12D, tion 80.89%. Pl.Exs. 13B. proposed

Voters within the district also Proposed House District 51 retains supported Michael Thurmond in the 1998 benchmark District 56’s status as a district 61.18%, primary runoff at a rate of and at majority with a population of total black BVAP, general 87.97% his election contest. Id. as shown below. (Ga.)

_TPOP25_BPOP_TVAP26 BVAP 41,311 32,393 17,724 Benchmark H.D. 56 24.801 _(60.03%)_(54.72%) 43,675 25,162 Proposed 34,793 18,118 H.D. 51 (57.61%)(52.07%) 12C, Moreover, proposed cycles PLExs. 13B. Dis- tion in the benchmark House Dis- trict 51 majority reg- 120, retains a black voter trict registra- 52.14% black voter 12D, istration level of 52.68%.PLExs. 13B. tion 52.90% in and 52.07% in Proposed House District 51 retains this HE, 2000. PLExs. 14B. making up status while benchmark District Voters within benchmark House District significant population 56’s shortage from 120 tend to vote for Party Democratic 45,480.

the current ideal district size of candidates as demonstrated an overall 1,805 Proposed District persons, 51 is 62.80%, Democratic Performance score of 3.97%, or short of the ideal district size. as well performance Democratic 12C, PLExs. 13B. numbers for the years individual election Proposed b. House District 95 (66.77%) (61.37%), of 1996 and 2000 Proposed single- House District 95 is a (57.06%). HE, Specifically, PLExs. 14B. member district includes all of Han- 77.32% of the voters within House District cock, Glascock, Taliaferro and Warren 120 voted for Thurmond the 1998 Demo- Counties, Baldwin, parts McDuffie cratic Party primary runoff In election. Counties, and Putnam Georgia. PLExs. addition, 72.74% of the voters House 12A, 14A. district includes District 120 voted for Thurmond in his territory currently within benchmark general 11E, victory. election PLExs. House District encompasses which all 14B. Voters within House District 120 Taliaferro, Warren, Glascock and Han- supported have also other African Ameri- cock parts Counties and of McDuffie and can Democratic running candidates Baldwin Counties. PLEx. 11A. statewide office. 1998 Thurbert Baker

The benchmark received House District 120 64.49% of the district’s votes 7,056 15.51%, people, and, Attorney short of his race for the ideal General 11D, district size. PLExs. Burgess 14B. Black vot- David garnered 72.17% of the er registration numbers have remained vote for Public Service Commissioner. Pl. relatively steady HE, over past three elec- Exs. 14B. population. "TPOP” refers to total voting age popula- 26. "TVAP” refers to total tion. PLEx. 14B. In using 2000 results. cur- 51.97% 120 is House District

Benchmark Hud- Representative performance Sistie past political at the rently looking held at 4. In Int.Ex. 31 son, Democrat. a white territory within contained pri- faced a Hudson Representative pro- within the House District an African challenge from mary supported Michael posed district also 4-5. Favors. Id. at Frederick opponent, a rate of 75.01% the 1998 Thurmond at received 51.3% Hudson Representative runoff, general in his primary and 67.36% 48.7%. vote, Favors received while Ms. Additionally, these vot- election contest. Id. Baker at a rate of supported Thurbert ers an overall Demo- calculated Plaintiff has PL Burgess David at 66.57%. 59.78% and of 56.61% for score performance cratic 12D, 14B. Exs. addition, In District 95. proposed House of benchmark Dis- demographics The performance projects Democratic the State 95 Proposed House District trict 120 and 1996 election re- using of 53.04% numbers results, are shown below: sults, and using 1998 62.65% past 117 over the 12C, existing House District 11D, Additionally, 14B. PLExs. cycles, from 71.80% pro- in the three election registered voters 44.06% of finally to 75.59% in American. 95 are African to 75.32% posed District HE, 15B. 12D, 2000. PLExs. 14B. PLExs. up makes benchmark

This district also the benchmark House within Voters from the population shortage District 120’s heavily for candi- 117 tend to vote 45,480. Pro- size of current ideal district Party, demon- Democratic dates of the persons, District 95 is posed Democratic Perfor- by an overall strated Pl. 1.96%, district size. of the ideal short 79.13%, well as the mance score 12C, 14B. Exs. numbers for the performance Democratic (76.11%), years of 1996 individual election Proposed District 97 c. House (79.93%). (77.73%) PLExs. single- House District 97 is Proposed HE, of the voters Specifically, 73.73% 15B. wholly within district located member 117 voted existing House District within 12A, Georgia. PLExs. County, Richmond Demo- in the 1998 for Michael Thurmond territory embraces House District 97 15A. election, and runoff Party primary cratic House District included within benchmark general elec- for him 85.27% voted within wholly also contained which is HE, Additionally, vot- PLExs. 15B. tion. 11A. County. PLEx. Richmond House District within benchmark ers support for other electoral demonstrated 117 is House District The benchmark candidates Democratic African American 24.01%, 10,918 short of the ideal people, or voting at a office 11D, running for statewide vot- 15B. Black size. PLExs. district Baker for Thurbert rate of 83.09% increased levels have registration er Burgess. primary Thurmond in the runoff and 85.47% David PLExs. HE, 15B. supported him in elec- general 70.03% Additionally, tion. 67.74% these voters

Proposed District 97 retains much of the supported Thurbert Baker and 71.11% performance overall Democratic of bench- 12D, supported Burgess. David PLExs. mark District 117. Plaintiff has calculated performance an overall Democratic score 15B. District, of 68.59% for this House District 97 also

projected performance Democratic num- retains District 117’s status as a district results, using bers of 60.47% 1996 election comprise which African using 64.37% 1998 results and 63.08% us- majority voting age of both the total and ing 2000 election results. Pl.Ex. 15B. populations. following compares table looking past political performance at the demographics of the benchmark Dis- territory proposed contained within 97, proposed 97, trict 117 and House District House District 72.92% of the voters in proposed supported using figures: district Michael the 2000 census 11D, 12C, Additionally, pro- PLExs. 15B. tion have increased in the House existing posed District 97 retains District ma- past 117’s District 116 over the three election jority registration black voter at a level of cycles, from 52.75% of total registra- voter 12D, 1996, 54.53%.PLExs. No 15B. 1998, tion to 54.59% and to HE, 54.29% 2000. PLExs. 16B. 1,949 Proposed per- District 97 is sons, 4.29%, or short of the ideal district Voters within benchmark House District 12C, size. PLExs. 15B. tend heavily to vote for Democratic candidates, Party as demonstrated an Proposed d. House District 100 performance overall Democratic score of

Proposed House District 100 single- is a 62.13%, performance and the Democratic member district that contains all of Burke years rates for the individual election County part and County, Richmond (62.88%) (60.72%), and 2000 12A, Georgia. proposed PLExs. 16A. The (59.83%). HE, PLExs. 16B. Specifically, district territory 100 embraces included 74.88% of the voters benchmark House within benchmark House District District 116 for voted Michael Thurmond which also contains all of County Burke Party primary his 1998 Democratic run- part County, and of Richmond Georgia. election, supported off and him in 69.96% PLEx. 11A. HE, general election. PLExs. 16B. Furthermore, population

Based on the voters in benchmark statistics from House census, supported the 2000 District 116 have other benchmark House African 6,161 13.55%, persons, District 116 running American candidates statewide office, short of the ideal voting district size. PLExs. at a rate Thur- of 69.05%for 11D, 16B. Levels of black voter registra- bert Baker in and for David 74.72% HE, runoff, primary PLExs. 16B. Pro- mond in the 1998 Burgess supported him in general 70.37% his elec- 100 retains much of posed House District tion PLEx. Additionally, contest. 12D. performance Democratic the overall supported 69.48% of these voters Thurbert projected with a district benchmark supported Baker and 74.68% David Bur- performance score of overall Democratic 12D, gess. PLExs. 16B. 62.56%, perfor- projected Democratic 1996 election using mance rates of 61.13% Like benchmark proposed results, results, using 1998 63.53% House District 100 has a majority African PI. for the 2000 election results. 60.14% population age popu- past political looking Ex. 16B. In following compares lation. The table territory contained performance demographics of the benchmark District House District within 116 and House District us- supported Michael Thur- ing figures: 74.98% of voters 2000 census 11D, 12C, 1996, 1998, Additionally, pro- tion in finally PLExs. 16B. to 74.26% and HE, to 76.34%in 2000. PLExs. 17B. posed majority House District 100 has a registration black voter level of 54.67%. Voters within the benchmark House Dis- 12D, PLExs. 16B. heavily trict to vote 134 tended Demo- candidates, Party cratic as demonstrated Proposed House District 100 is by an overall Democratic Performance 1,287 2.83%, persons, or short of the ideal 84.72%, per- score of and the Democratic 12C, size. PLExs. 16B. district formance rates for the individual election (87.93%) (68.06%), years of 1996 and Proposed e. House District (84.58%). HE, PLExs. 17B. Specifi- Proposed single- House District 113 is a cally, of the 71.70% voters bench- wholly member district located within mark District 134 for Michael House voted 12A, Muscogee County, PLExs. Georgia. Party Thurmond the 1998 Democratic proposed 17A. The district embraces terri- election, primary runoff and 90.11% of vot- tory included within benchmark House for him in the general ers cast their ballots 134, wholly District which is also contained HE, Additionally, 17B. election. PLExs. County. 11A. Muscogee within PLEx. within the benchmark House Dis- voters significant trict 134 demonstrated electoral population Based on the statistics support for other African American Demo- census, the benchmark House running cratic candidates for statewide of- 31.92%, 14,518 persons, District 134 is supported of voters fice. 91.03% of the ideal district size. PLExs. short and, Thurbert Baker 88.90%voted 11D, registration 17B. Black voter levels HE, 17B. Burgess. for David PLExs. existing have increased in the House Dis- past cy- trict election Proposed 134 over three House District 113 retains cles, perfor- much of Democratic registra- from 72.87% of total voter the overall supported general him in elec- District with an 77.38% his

manee of benchmark Additionally, performance tion contest. 79.45% of these overall Democratic score Baker, 75.04%, projected supported Democratic Thurbert specific Burgess. Pl.Exs. using supported of 64.81% 78.56% David performance numbers results, 12D, using 75.24% 17B. 1996 election results, using and 73.16% 2000 election 134, proposed Like benchmark District results. Pl.Ex. 17B. majority of African District 113 has political performance population voting age popu- past

Based on the compares the territory proposed following contained in lation. The table District demographics House District 69.38% of voters within the benchmark us- supported district Michael 134 and House primary ing figures: in the 1998 runoff and the 2000 census Thurmond 11D, 12C, Additionally, pro- 17B. Pl.Exs. included benchmark House Districts 148 posed majority wholly House District 113 has which were also contained registration at a County. black voter level within Chatham PLEx. 11A. 12D, 61.88%.PLExs. 17B. According population statistics Proposed only House District 113 is census, the 2000 Dis- benchmark House 1,674 3.68%, persons, or short of the ideal 10,343 persons, trict 148 was short 12C, district size. PLExs. 17B. 22.74%, and benchmark 149 District was 12,815 28.18%, persons, or short of the plan, proposed Under the two incumbent 11D, ideal district size. PLExs. 18B. Black minority legislators, Representatives Caro- in registration voter levels have increased lyn Hugley Taylor, and Maretta are drawn the benchmark District House 148 over proposed into the District 113. See Int. past cycles, three election from 64.19% (Westmoreland Decl.). Ex. 31 8 Inter- 1996, 1998, in in finally to 65.37% and testimony venors introduced that these are HE, in 68.12% 2000. PLExs. 18B. Black two incumbent Democrats includ- registration voter levels have decreased in any proposed any ed in one in district benchmark House District 149 over the plans the three before court. Id. An this past cycles, three election from 70.80% of open proposed seat was created in House 1996, registration total voter in to 70.46% 109, majority-white adja- District district 1998, in finally to 67.96% in 2000. and. proposed cent to district 113. Id. HE, PLExs. 18B. Proposed f. House District 124 in benchmark 148 Voters House District Proposed 124 Party House District is a two- tend to vote for Democratic candi- dates, wholly by an member district located within as demonstrated overall Dem- 76.58%, County, Georgia. performance PLEx. 12A. Chatham ocratic score and Proposed 124 territory performance District embraces Democratic numbers for indi-

53 Baker, and, (72.60%), in for bert 85.90% voted 1996 1998 years of election vidual (74.75%). HE, (79.26%) HE, Pl.Exs. 2000 PLExs. 18B. Burgess. and David in of voters 82.23% Specifically, 18B. 124 Proposed House District has an for 148 voted House District benchmark performance score of overall Democratic in the 1998 Democratic Thurmond Michael 67.67%, projected perfor- Democratic election, and 83.02% runoff Party primary using mance numbers of 63.16% 1996 elec- Pl. election. general him in the supported 69.79%, results, results, using tion HE, Additionally, voters within 18B. Exs. election results. and 66.46% for the 2000 148 demon- House District benchmark looking past political In at the PLEx. 18B. for other African support strated electoral territory contained running performance candidates of the Democratic American In 82.90% statewide office. proposed for House District within At- Thurbert Baker for supported proposed of voters within the dis- 81.50% and, sup- 82.83% torney General supported Michael Thurmond trict HE, PLExs. 18B. ported Burgess. David supported primary runoff 75.83% House District him in election contest. 76.11% general from benchmark his Voters Party support tend to Democratic also Baker supported of these voters Thurbert candidates, by an overall as demonstrated supported Burgess. Pl. and 75.39% David 81.20%, score of performance Democratic 12D, Exs. 18B. performance numbers and Democratic Districts 148 and As benchmark (77.61%), years of 1996 election

individual District African (78.35%). (81.89%) PLExs. and 2000 up majority the total voters make both HE, of the voters bench- 18B. 89.39% populations. following voting age voted for Michael mark House District 149 compares demographics table runoff, primary Thurmond in the House Districts 148 and benchmark general him in the elec- supported 88.84% 124, using House District 89.34% of voters PLEx. 18B. tion. figures: Thur- the 2000 census 149 voted for benchmark 4.60%, size. Pl. 12C, of the ideal district 11D, Additionally, pro- short 18B. PLExs. 12C, majority Exs. 18B. 124 has a black posed District of 55.21%. PLExs. registration level voter

12D, Proposed 18B. g. House District 125 single- 125 is a proposed Proposed Dis- House District district size for

The ideal wholly within 90,960, district located the ideal district member 124 is twice trict 12A, Chatham, 19A. Thus, Georgia. PLExs. district. single-member of a size territory in- 125 embraces 2,091 House District persons, or support electoral for other eluded within benchmark House District demonstrated wholly within African American Democratic candidates which is also Chatham in running Pl.Ex. for statewide office: County. 11A. and, in voted for Thurbert Baker 72.36% population Based on statistics from the Pl. Burgess. 72.78% voted for David census, the benchmark House Dis- HE, Exs. 19B. 11,450 25.18%, persons, trict 151 is Proposed an of the ideal district size. PLExs. House District 125 has short 11D, performance Black voter overall Democratic score of registration 19B. levels 59.40%, specific projected in House Dis- Democratic have increased benchmark past cy- performance using trict 151 over the three election levels 56.64% cles, results, results, in in using from 46.18% 1996 to 47.93% election 62.75% 11E, using and to 48.40% in 2000. Pl.Exs. and 56.56% 2000 election results. pro- 19B. Pl.Ex. 19B. 70.77% of voters posed supported Thur- district Michael Voters benchmark House District 151 runoff, primary mond the 1998 Party tend to for Democratic candi- vote supported him in general 67.13% his elec- dates, as demonstrated an overall Dem- Additionally, tion contest. 67.84% of these 64.31%, performance ocratic score of Baker, supported voters Thurbert performance the Democratic numbers for supported Burgess. 68.40% David PLExs. years individual election of 1996 12D, 19B. (67.19%) (61.39%). (61.47%), 1998 and 2000 HE, Specifically, Proposed PLExs. 19B. 75.40% of While House District 125 has majority population, benchmark House District 151 a it does not majority voted for Michael Thurmond the 1998 have a The following BVAP. table election, Party primary compares demographics Democratic runoff of the bench- him supported and 71.67% voters the mark House District 151 and HE, general election. PLExs. 19B. Vot- House District using the 2000 census figures: ers benchmark House District 151 have 12C, tricts. See PLExs. 11D, Proposed 12C,

PLExs. 19B. House 12D. Plaintiffs registration District 125 has a voter expert report districts, black lists the 12D, level of 41.67%. PLExs. 19B. Pro- and the number of seats and of BVAP each posed only persons, District 125 is or district. Pl.Ex. pro- 25. There are six 1.84%, short of the ideal Pl. district size. posed multi-member districts with BVAPs 12C, Exs. 19B. higher: of 50% or District 43 has BVAP 65.18%; of District 48 has a BVAP of h. Multi-Member Districts 61.13%; 61.86%; District 49 has a BVAP of 59.51%; Some of in Georgia’s pro- the districts District 60 has a BVAP of posed 58.33%; plan are House multi-member dis- 61 has a BVAP of and District 124 3; pre- Redistricting the Office did not App. Pl.Ex. of 54.14%.27 has a BVAP According to 2/5/02, political per- Tr., pare demographic a.m. at 79-81. that mi- probability expert, plaintiffs plan reports formance SENPLAN2U their candidates of will elect nority voters Friday it morning until the of the was certainty 80% to near from choice varies 702, 76:6-77:22. enacted. U.S.Ex. . Id. districts. multi-member these reappor- The second Senate Bill 1EX1 tionment became Senate Reapportionment Plan F. Senate State Georgia adopted and was State mandates Georgia Constitution 10, 2001, by a vote of 29 August on Senate of 56 Senators consist Georgia Senate ¶¶ 412, Regina PPFF 413. Senator to 26. ap- single-member districts elected only area was the Thomas of the Savannah respective districts among portioned African American Senator who voted Const., Ill, Art. Georgia. Ga. the State ¶ plan. Id. 413. The House of against ¶II, 1(a); § Members § 28-2-2. O.C.G.A. Bill Representatives passed 1EX1 Senate are elected for two- Georgia Senate of the 17, 2001, August by a vote of 101 to 71. until the time of the and serve year terms Pelote, Representative Dorothy also of Id. Assembly. of the next General convening KV(a). area, was the African Ill, II, the Savannah Const., § Mem- Art. Ga. representative against are elected to vote Georgia of the Senate bers Ga. plan. signed same time as Governor. Id. The Governor Senate ¶I, Const., V, § II. 24, 2001, Art. August Bill 1EX1 into law on ¶ Id. 414. Act No. 1EX6. a state- 2000 census results show 8,186,453 people. Con- population wide for the 56 Senate the ideal size

sequently, Demographics 1. of Benchmark person-one' purposes of one Districts for Proposed Districts ¶ 146,187 PPFF 407. The people. vote is current Georgia’s The State of Senate have Georgia’s districts Senate State of a mediated districting map is the result a deviation traditionally been drawn with agreement between State from the percent minus five plus Department of Jus- and the United States ¶ According to size. Id. 411. ideal district the 2000 census upon tice in Based data, only 2000 census two statistics, thirteen Senate Dis- there are districts with a total Senate benchmark plan with a tricts in the benchmark BVAP, or black vot- population, total PPFF of over 50%. population total black 50% were within that registration er over ¶ 2, 10, Districts Pl.Exs. at 408. These requirement. traditional deviation 38, 39, 43, ID, majority 12, 15, 22, 26, 35, 36, ten 44 and IE. All three of the but ideal deviated from the BVAP districts Districts The same thirteen Senate Id. Id. by -14.02% to -26.94%. district size levels of registration total black voter ¶ 410. Twelve higher. PPFF at 50% 10, 2001, Redis- the Senate August On have BVAP of 50% or of these districts leadership called for tricting Committee ¶ Dis- Id. 409. These are Senate higher. redistricting final on a Senate vote 38, 39, 10, 12, 15, 22, 26, tricts SENPLAN2U, after shortly plan, cur- District which and 55. Id. Senate distributed to Sena- plan had been population total black rently has 52.80% Meggers Ms. testified tors for review. calculating BVAP. Georgia’s method 27. BVAPnumbers reflect 56 is When BVAP jority-minority BVAPs. registration, has voter

and 53.72% black Attorney with the calculated accordance BVAP.28 49.62% Guidance, Dis- proposed Senate General’s proposed plan, Like the benchmark 2 and 34 fall below 50% BVAP. tricts Dis- plan contains Senate redistricting Furthermore, population of the black population of over with a total black tricts severely re- 44 would be Senate District However, proposed plan con- 50%. plan. proposed Bench- duced under in which eight Districts only tains almost 44 has a BVAP of mark District The registration is over 50%. black voter 44 has a proposed District 50%. plan also proposed contends that State approximately 34%. BVAP of in which the creates 13 Senate Districts compares following table the demo- 50%. The United States BVAP is over Dis- of the benchmark Senate According graphics to the this calculation. disputes Districts, census tricts with interpretation of the United States’ the 2000 census statistics: data, using with ma- there are districts 35,629 110; District 2 was PLEx. 25. of benchmark Senate U.S.Ex. 24.37%, or ideal size. people, short Districts Contested ¶ 428; PLEx. 1-D. The PPFF see also a. Senate District 7,217 people, 2 is proposed District Sen- Both the benchmark 4.94%, ideal district Id. short of the size. within Chatham ate District are located ¶ 429. the ideal district County, Georgia. Given statistics, According to 2000 census 146,187 persons for a Senate dis- size of census, population of the benchmark Senate trict after the 2000 BVAP Guidance, according General’s the BVAP of District BVAP is calculated This method; Georgia's according Attorney to the 48.52%. op- representing Senate District without Pl.Ex. 3B. 60%. approximately ¶ in the General Election. Id. 435. position 2 is District Senate BVAP 50.31%, plaintiff, according either 1999, was In Senator Johnson indicted States. 49.81%, according to the United of federal mail fraud. Id. and convicted argues ID; 110. The State Pl.Ex. U.S.Ex. ¶ from the resigned 436. She State Senate alone, that, changes population because 12, 1999, November Governor of District that the BVAP it was inevitable fill vacant special election to called 2 would decrease. special seat. Id. This election was Senate by Department of Justice. precleared compose 48.42% registered Black ¶ (DOJ (Dec. 437; File No.1999-3631 2. In con- Id. District proposed Senate 1999)). following quali- candidates at least six District trast Senate special nonpartisan run in the elec- districts, fied to reg- majority-minority 21, 1999, to succeed tion held on December than BVAP. The higher levels istration Pro-Life Anderson Johnson: calculations of Senator suggests States United (white man), (white man), F. Braun Dana include an may for Senate District BVAP (African man), Bob Brown American Willie ineligible voters. number of undetermined (African man), American Charles Bryant according Meg- to Ms. District Senate (African man), Regi- American Universi- Gordon State gers, is home Savannah woman). (African Thomas of Art and na School ty and the Savannah ¶ Id. Thomas won that election. Regina was Id. Meggers 177. Ms. Design. USPFF at the the number of students unaware of was a Thomas testified she Senator schools, on-cam- the Schools have whether Reapportionment member of Senate the students or whether all of pus housing, However, she testified Committee. further voters. Id. She registered were process was redistricting her role 1,500, population that a student testified did the committee” and she “[jjust to be on “large,” not consider which she would every more than attend anything not do 1.5% of aрproximately would constitute 704, 38:25-39:3, 40:12- meeting. U.S.Ex. population. voting age 2’s District Senate redistricting against 24. She voted Id. expressed displeasure her plan and has Id. of her district. Thomas, re-drawing with the an African Regina Senator difficulty that she will American, She believes incumbent Sen- is the current and that African Ameri- being re-elected 2 has been 2. District ate District opportunities to have fewer can voters will African American by several represented choice the re-drawn Al- candidates of Roy L. elect opposition, Senators. Without attempts to dis- 2. Plaintiff American, District len, won the Senate is African who “surpris- being Thomas as credit Senator and the Gen- Party nomination Democratic ¶ demographics of with the ingly unfamiliar PPFF 431. With- Election eral indicated at her Johnson, she is her district” because Diane H. who opposition, out her district had testimony that American, deposition won the Democratic African ¶ BVAP, PPFF at 421 not 50%. 49% Election nomination and the General Party 146^17). ¶¶ Dep. at Thomas (citing In Senator PPFF in 1994 and 1996. above, However, using as discussed opposed was Johnson Senator mandated calculating BVAP pri- method in the Democratic Regina Thomas Guidance, proposed United States’ by a 297-vote mar- won mary, but Johnson ¶ than a BVAP of less Id. 434. Senator gin victory. 50%. to her third term was re-elected Johnson

58 Shinhoster, addition, an Richard Ex- election, Thom- runoff Senator

In a 1999 of the local member ecutive Committee opponent, Dana her as defeated white NAACP, majority testified a slim that votes,29 Braun, than 70 by slightly more a sufficient to elect may BVAP not be 2 had a District when benchmark Senate if voter turnout candidate of choice rates more than 60%. level of registration low; community the African American are ¶ in No- Thomas ran PPFF 440. When that a black can always “cannot be assured reeleetion she vember won majority elected when —when 22,723 a by Election vote General ” dep. ratio is so close.... Shinhoster at ¶ (22.2%). (77.8%) 6,494 Her Id. claims that an in- 16. Shinhoster also a general election was opponent that crease BVAP in Dis- apparently did not Republican, who white “voting pattern[s] trict is needed because Republican Party. support of have the po- are County and Chatham Savannah (Senate mi- See U.S.Ex. 86:19-88:13 ¶509 8 along larized racial lines.” U.S.Ex. Republican Party de- nority leader for (Shinhoster deck). hand, on On the other a “nut case” opponent Thomas’ scribing cross-examination, City Savannah Aider- that Republican “flaky, and half-crazed “people testified that man Helen Johnson then”); see also runs for office now and of racial issues or really don’t vote because 187:23-188:2, 704, 32:19-23, U.S.Ex. particular right race in the because their (Thomas opponent 189:19-25 testified dep. at 41. city elections.” Johnson man, delivery not well known pizza was a lay witnesses from Senate Several had been arrested community, and 2 presented District the United States a officer and car- impersonating police Tybee addition of the testified that Although Sen- rying weapons). concealed Island, Hope, and Thunderbolt ar- Isle of that 2000 election ator Thomas won District “will eas majority, did not receive a substantial she Ameri- make it more for African difficult majority of the white vote. U.S.Ex. can voters to their chosen candi- elect ¶ (Shinhoster presented testimony The United States dates.” U.S.Ex. 509 ¶ (D. deck). deck); from Chatham Jones from eleven witnesses U.S.Ex. County proposed Senate about Despite polariza- the existence of racial Senators, four African including two State tion, testified that Af- these witnesses also Aldermen, two City Council of choice can be rican American candidates Commissioners, County Chatham they professional elected to office when three members of the Executive Commit- well-respected by the white communi- Branch of tee of the Savannah example, the United ty. For some of Almost all of the witnesses testi- NAACP. they States’ witnesses stated consid- polarized racially fied to the existence of Regina strong ered to be a Afri- Thomas example, 2. For Senate District popular very can American candidate and Jackson, Dr. Prince Executive Commit- an dep. in her Jackson Senate District. at NAACP, 62-63; 26-28, tee member of the local testified dep. Shinhoster 68. Al- it African-American is his “belief that Gwendolyn derman Goodman testified that, typically despite tendency people voters vote for African American vote candidates, race, typically vote according an African Ameri- when white ¶ professional 503 8 for white U.S.Ex. can candidate is known as candidates.” deck). (Jackson respected by community, the white margin victory between 73 and 76 votes. 29. The was “go proposed will with what is contrast to the many white voters Senate District 12, dep. registration Goodman at 29-80. black are right.” higher levels in than at least six of the BVAP other expert, Dr. Richard The United States’ proposed majority-minority districts. The local election data Engstrom, analyzed suggests United States that the BVAP in 2, in- the benchmark District Senate proposed 12 may Senate District include municipali- in cluding data from elections an ineligible undetermined number of vot- ties and counties within benchmark Senate ers. Senate District 12 includes or two specific results of as- The his colleges Albany, three small in and state racially polarized voting are sessment prison 733, in County. Calhoun U.S.Ex. found in Part 2G of this Memorandum Meggers 124:22-23. did not know the en- Opinion. colleges, rollment of the but testified that b. Senate District prison population approxi- inmate was 1,100. Furthermore, .mately Id. John proposed Both the benchmark and Sen- signifi- testified to his belief that a White in ate District are located southwest cant of African number American felons Georgia. The benchmark Senate District in ex-felons live the district and cannot 25,982, 17.8%, 12 is below the ideal law, reducing vote under state thus According district size. to 2000 census eligible voting age African number data, the benchmark Senate District existing proposed Americans in (Ga.) has a BVAP of 55.4% or 54.94% ¶ district. U.S.Ex. 513 27. (U.S.). before the Under Senate court, 12 will have a Senate District BVAP past voting performance pre- (U.S.). (Ga.) of 50.66% or 50.22% Plaintiff proposed cincts contained within Senate argues drop that some in BVAP was inev- past District 12 in certain statewide elec- ¶ 4B; PLEx. PPFF 474. Howev- itable. pro- tions shows that 58.57% of voters in er, majority-black precincts were several posed supported District 12 Mi- Senate removed from benchmark Senate District run-off, primary chael Thurmond in his in drawing proposed district. supported while 65.89% of voters him in U.S.Ex. 118. 1998, general élection. PLEx. 4B. In voters in BVAP, 64.53% of the Senate reducing pro-

In addition District 12 Baker for voted Thurbert posed District 12 reduces the black and, Attorney 65.83%vot- General registration percentage as a of overall reg- Burgess ed for David for Public Service approximately istration from 52.5% to PLEx. 4B. Commissioner. In registered 47.5%. 48.16% оf the the benchmark Senate District 12 Primary, In the 1992 Democratic Mark ¶ were African American. PPFF 483. Taylor, a white man and current grew That number to 50.69% 1998 and Governor, Charles Lieutenant defeated Using Id. to 52.48% 2000. man, Sherrod, an African American to ob- lines, percentage Senate District nomination for Dis- party tain registered voters would have been ¶ Taylor 12. PPFF trict 490. Mark won in 1998 44.86% 46.57% and 47.56% 12 in the 1992 election to Senate District Id.; in 2000. Pl.Ex. 4B. opposition. Election without Id. General percentage registered Taylor unopposed voters who In was black, 47.46%, Primary Elec- is less that the BVAP of Democratic and the General Taylor and won re-election. Id. Mark proposed Senate District either 50.66% tion ¶487. (U.S.). (Ga.), that, PPFF In when first elected to the or 50.22% testified about Taylor won with that “Mr. stated Senate, political endorsement he had badly very I did Taylor of the votes. community. two-thirds African American ¶ White, an Afri- 510 6 community.” U.S.Ex. dep. at 20. John in the white candidate, Taylor deck). opposed (Sherrod explains can American He “whites ” PPFF Party primary. my opponent.... the Democratic vote for white would ¶ by margin defeated White Taylor however, deposition, in his admits Sherrod (38.5%) (61.5%) 9,406 votes. 15,043 to Tay- for voters also voted that rural black general to win the Taylor went on Id. Taylor’s financial resources lor and that Id. opposition. election without dep. at his success. Sherrod greatly aided White, that acknowledges 58-59. Sherrod in ran for Lieutenant Governor Taylor 1996, competitive campaign ran a more 12 in seat be- and the Senate District 1992; ¶ he did in White Taylor ran and lost than open. against Id. 492. White came attorney, financing, organization, Michael better a local white narrowly to had better Bremen, in the Democratic announcer on Meyer popularity von as “an as well as Bremen primary. Meyer Id. von Party the TV.” Id. of the votes. primary by 51.4%

won the that “most in his affidavit Sherrod states expert report demon- Engstrom’s Id. Georgia simply in voters Southwest white Party pri- that in that Democratic strates for black Senate Candidates.” will not vote voted virtually all the black voters mary, ¶ opinion 11. He bases his U.S.Ex. White, of the white and almost none for get sup- more assumption that “whites 601, Ta- U.S.Ex. voters voted White. get than port black[s]” from black voters 2. ble voters, that if the but indicated from white won Meyer subsequently Bremen von true, change it would his reverse were in won re-elec- general election dep. at 94-96. opinion. Sherrod the current incumbent tion is A Uni- in District 12. Id. Clark Senator Arthur City Albany Commissioner Meyer von Bremen to versity study found “I believe Mr. White Williams testified: minority issues. Pl.Ex. supportive primary District 12 1996 Senate [the lost However, experts for the 23 at 9-10. both compete he could not election] because have found plaintiff and United States money and Taylor’s Mr. because not the Meyer von Bremen is that Senator Taylor’s Mr. use responded to white African American candidate of choice of of the race card.” U.S.Ex. 511. Williams 2; App. U.S.Ex. voters. Pl.Ex. Taylor played that the “race contends Furthermore, of the Unit- several Table campaign by sending out during card” his witnesses testified Senator ed States’ compared White Williams literature represent Meyer von Bremen does to vote in the Demo- urged citizens the African American commu- interests of White, Party primary against and to cratic ¶ deck); (Wright nity. U.S.Ex. Republican Party primary. not vote ¶ (D.Williams). U.S.Ex. 512 ¶¶ 6, suggests Plaintiff that the com- Id. Taylor’s parison of to Williams White In addition to the statistical evidence ap- was not a racial campaign literature racially polar- existence of suggesting the fact peal, explained by rather is but voting District the Unit- ized posi- and Williams similar White had witnesses testified about their ed States’ concerning municipal con- tions on issues racially polarized voting. experiences of minority-owned businesses and tracts with Taylor, Mark Describing his 1992 defeat to employment. candidate, city action Mr. affirmative a white Charles Sherrod ¶ (White ¶ expert analyzed The United States’ local 518 9 (citing U.S.Ex. PPFF 77-78). election data from the benchmark Senate decl.); dep. A. Williams specific 12. The results of his defeat at that White’s argues The State polarized racially assessment of racially polarized polls was not due in Part 2G of this Memorandum found nega- the result of was rather voting, but Opinion. statewide in the local and tive articles the Atlan- published press. One article c. Senate District 26 28, 1990, stated: ta Constitution on June Both the benchmark and Sen- something appealing about “There is substantially ate District 26 are located Representative of State naked self-interest County, Georgia. According within Bibb public service. approach John "White’s data, to the 2000 census the benchmark all that embrace of bad His blatant underpopulated by District 26 is any to be offensive to politics ought about 42,119 people, or is 28.8% short of the ideal hpw voter, you get angry can at such but district size. The benchmark Senate Dis- 28-29, Ex- dep. at greed?” candid White (Ga.) trict 26 has a BVAP of 62.45% or that such articles 1. White admits hibit (U.S.). Proposed 61.93% Senate District voters’ decisions might have affected white (Ga.) 26 has a BVAP of 50.80% or 50.39% opponent. him his White to vote for (U.S.). 6-7,18. dep. at percentage registered of black vot- may that he not be a stated Wdiite ers in the benchmark District 26 is 62.79%. among white voters. strong candidate ¶ percentage PPFF That decreases 529. 29, dep. at He testified that White District 26. Id. proposed in the to 48.27% candidates were three African American registered voters in The number black getting at white votes more successful grew the benchmark Senate District 26 he, explained than but African was to 48.27% from 1998 to the from 46.51% to garner American candidates were able ¶ Id. 531. Levels of black year 2000. support of they where had the white votes precincts compris- in the registration voter at figures. dep. WTiite 34- political white proposed District 26 increased ing Senate 35, 38, that he lost 123-24. White believes registration total voter in from 46.5% of Meyer von Bremen because “some black year in the 2000. PbEx. to 48.3% they fig- turn people did not out because 5B. to win.” ured was sure bet [White] District proposed In contrast to Senate ¶ deck). (White U.S.Ex. 514 15 higher levels are registration proposed in at least six of the of the United States’ than BVAP Finally, several majority-minority districts. The United that it would be close witnesses believed in Dis- suggests candi- States BVAP Senate impossible for an African American includes an undetermined number proposed in the Senate trict 26 date to be elected ¶4 in ineligible residing voters the district. (Wright District U.S.Ex. ¶ ¶ (D. deck). 382; 733,126:2-22. Ms. deck); U.S.Ex. Williams USPFF U.S.Ex. witnesses, however, University testified that Mercer Meggers stated Some of these proposed in District is located Senate supported that white voters have some the full- that she does not know candidates such as an but African American university. time student enrollment of Superior Judge, Court African American Id. testified that he Michael Brown also Bishop and Senator Congressman Sanford prison popula- 25-26, ineligible dep. at 83-84. did not consider Wright Thurmond. won a seat on the Bibb opponent and Dis- white Senate drafting tions in ¶ when County Board of Education 38B. trict 26. USPFF per- and the County’s BVAP was 43% past returns for According precinct voters was registered of black centage elections, precincts in the 71.67% of voters 13:14, 11:13. dep. at 39.86%. Abrams 26 voted that he acknowledged benefited Abrams Thurmond his 1998 Michael 71.67% for Id. at 21. voting. crossover from white voted for Mr. primary runoff 68.58% Hart, District Commissioner F. Samuel Election Thurmond in his 1998 General County Board vice-chair of Bibb 2D, 7B. In 66.92% of race. Pl.Exs. Commissioners, that he received testified precincts voted for same those portion of the white vote. substantial Attorney in his General Thurbert Baker 24. Board of dep. Hart Education race, Burgess David and 68.91% voted Barnes also testified that Member William Commissioner in his 2000 Public Service has been suc- Senator Brown he believed 2D, 7B. race. Pl.Exs. Barnes winning white votes. cessful current incumbent Brown is the Senator dep. at 54-55. original- was District 26. Brown States’ witnesses Several United special in a election ly elected they did not believe that testified Reichert, man, a white defeating Robert would be able to African American voters (56.71%) 5,651 7,403 of by margin choice other than the a candidate of elect ¶ (43.29%) PPFF 535. Senate Dis- votes. See, incumbent, Brown. current Senator had a total black trict 26 at that time ¶ (Bivins deck); e.g., 517 10 U.S.Ex. 46,623 (58.0%), and BVAP of population of ¶ (Hart 725, deck); 519 9 U.S.Ex. U.S.Ex. (52.79%). 31,350 registered 47.4% However, the witnesses 36:15-37:3. also Id. American. at that time were African of other viable Afri- affirmed the existence 1996 Democratic In the 1994 and dep. Abrams at can American candidates. elections, primaries general and in the 45, 61; dep. Hart at 44-45. and elected without Brown was nominated expert analyzed local The United States’ 2000 Democratic opposition. Id. In the from the benchmark Senate election data election, op- had no Party primary Brown specific 26. The results of his District election, general In the 2000 position. Id. racially polarized voting assessment Williams, Republi- Brown defeated Greg in of this Memorandum found Part 2G candidate, mar- by a can African American Opinion. 21,453 5,491 votes. Id. gin of stated Witnesses the United States d. 15 Senate District that, opinion, in their in Senate District is newly The Senate District 15 drawn for white candi- “most white voters vote County, wholly Muscogee within contained voters tend to vote dates and most black 2A, ¶ Georgia. PkExs. 5A. The benchmark U.S.Ex. 515 6 for black candidates.” encompassed all of District 15 Senate (Abrams 70; deck); dep. at Abrams County part of Musco- ¶¶ Chattahoochee (Bivins decl.); 9 U.S.Ex. U.S.Ex. 517 1A, County. 5A. gee ¶ PkExs. (Hart deck). 519 8 The Senate District 15 is African American candidates have been benchmark 39,385 26.94%, people, or short of the ideal attracting some white votes successful ¶ ID, 562; PkExs. 5B. States’ district size. PPFF County. Bibb United One 6,821 Abrams, witnesses, Proposed District 15 is against ran Senate Albert J. 63 runoff, for him primary and 68.97% voted 4.67%, ideal dis- short of the or persons, ¶ 2C, ¶469; PLExs. 5B. election. PPFF 567. Ad- general PPFF trict size. ditionally, sup- 71.18% of these voters 15 has a to- District Senate Benchmark Baker, and ported 70.34% voted Thurbert 65.75%, percentage of population tal black 2D, Burgess. PLExs. 5B. for David 60.93U.S.). 62.05Ga.) and BVAP ¶ Dis- proposed Senate 561. The PPFF Dis- The current incumbent Senate percentage of BVAP trict reduces Harbison, an Afri- trict 15 is Senator Ed (U.S.). (Ga.) PPFF or 50.05% 50.87% ¶ can man. PPFF American ¶ 568. Party primary Democratic for Senate vot- registered percentage Joseph Harbison defeated District Wi- in the bench- from 72.69% ers is reduced man, an African American for the ley, in the District 15 to 50.25% mark defeated party nomination. Id. Harbison ¶¶ 563, Black voter Id. District 15. Republican African Wright, his William increased numbers have registration opponent, General District 15 over the benchmark ¶ op- had Id. 572. Harbison no Election. cycles, from 69.19% past three election 1998 and 2000 position finally to in 1998 and 1996 to 70.79% primaries general elec- Democratic ¶ 563; IE, 5B. Id. Pl.Exs. 73.00% 2000. tions. Id. the benchmark Senate within Voters for Demo- heavily tend to vote

District 15 e. candidates, Party as demonstrated cratic newly drawn Senate District performance by an overall Democratic County, wholly within Richmond contained 81.59%, Democratic well as the score of ¶ 574; 2A, 6A. Georgia. PPFF PLExs. the individual elec- performance levels for 22 is also whol- Benchmark Senate District (82.93%) (71.23%), years tion ' County. ly contained within Richmond (81.79%). IE, 5B. 71.06% and 2000 PLExs. ¶ 575; 1A, PPFF PLExs. 6A. within the benchmark Senate of the voters in Thurmond 15 voted for Michael District 22 District is The benchmark Senate Party primary runoff 1998 Democratic from the ideal district size underpopulated election, him in the supported and 87.59% ¶ 580; 37,675 PPFF or 25.77%. people IE, 5B. Voters election. PLExs. general ID, Dis- proposed 6B. The Senate PLExs. District 15 demon- in benchmark Senate up shortage, 22 for this and is trict makes African support for other strated electoral 4.85%, below the ideal people, 7079 candidates, voting Democratic size. in 1998 at a rate of for Thurbert Baker 88.95%, Bur- 79.23% for David with 22 has a Senate District The benchmark IE, in PLExs. 5B. gess 2000. (Ga.). population of 66.84% total black ¶ Dis- proposed The Senate PPFF 579. an District 15 retains

Proposed Senate population of 22 a total black trict score of performance Democratic overall ¶ proposed 586. The Senate 54.71%. Id. 65.91%, perfor- projected Democratic experience 22 a decrease District would using 1996 election mance levels of 61.99% (Ga.) (U.S.), or 62.65% from 63.51% results, BVAP results, and using 1998 67.57% (U.S.). (Ga.) Id. or 50.76% to 51.15% results. Pl. using 2000 election 63.44% ¶¶ 576, reg- percentage of black 586. 2D, pre- 5B. 66.41% of voters Exs. fall from 64.07% would also istered voters comprising district cincts ¶¶ 581, Id. in the 1998 to 49.44%. Michael Thurmond supported 64 political previous trained and had some Senate within benchmark

Voters Dep. 81:19-25-82:3. experience.” Walker Democratic supported 22 have District candidates, by an as demonstrated Party Testimony Methodologies Expert on G. score of performance

overall Democratic Voting Predicting Patterns of num- 72.46%, performance Democratic and years of election bers for the individual single Georgia relied on a The State (71.34%), 2000 (71.02%), and attempt Epstein, expert, Dr. David (71.46%). IE, Specifically, Pl.Exs. 6B. proof on State meet its burden existing within 67.85% of voters House, and United States State Senate Thur- for Michael District voted Senate Epstein redistricting plans. Congressional Party pri- Democratic in the 1998 mond probit anal- single methodology, relies on election, sup- and 76.64% mary runoff all three as to ysis, for his conclusions election. PI. general him the ported expert, Dr. States' plans. The United IE, Additionally, of voters 6B. 73.97% Exs. on three varia- Engstrom, Richard relied District voted Senate the in- analysis, benchmark while regression tions of 77.41% Katz, Baker in and Thurbert lim- expert, for Dr. Jonathan tervenors’ in 2000. Pl.Exs. Burgess report critique for David to a scope ited the of his voted IE, analysis. The court Epstein’s probit 6B. scope Epstein and by the narrow struck 22 has an report ad- reports. Epstein’s Engstrom’s score of performance overall Democratic per- of what question dresses the limited 62.28%, perfor- projected Democratic produce 50% centage of BVAP will elec- using of 60.98% mance numbers an American candidate chance that African results, results, using 1998 63.19% tion Engstrom’s election. of choice will win 2000 election results. 59.98% degree report attempts ascertain ¶ 584; 2D, 6B. 65.09% of Pl.Exs. PPFF polarized voting existing racially pro- precincts comprising the report 26. Neither Districts Michael Thur- supported district posed question inquiry focuses its runoff, primary in the 1998 mond plans proposed redistricting whether the general elec- voted for him the 66.91% retrogressive. ¶ Additionally, these PPFF tion. by Baker supported Thurbert Analysis 1. Probit Id.; Burgess by 69.05%. and David 63.87% expert report probit describes Plaintiffs 2D, 6B. Pl.Exs. analysis as a statistical method standard is the current Charles

Senator Walker of an event determining the likelihood of State Senate District incumbent possible Epstein outcomes. that has two Majority in the Senate. is the Leader fitting a tool to describes the as method in to the Senate his election political Since to determine allow scientists only pri- one white Walker has Mr. faced that an African American candi- likelihood Moretz. USPP mary opponent, only David There are date of choice will win. ¶ Republican opposition possible 588. He has faced two with which the anal- outcomes man, time, by a D.L. John- candidate of ysis one white is concerned—the choice son, analysis Epstein’s Id. Sen- loses. is de- general in the 2000 election. wins or “point oppor- signed predict equal that he can win re- Walker believes ator election, tunity,” point that his successor or the which demo- but noted “well-financed, a result a 50% relatively graphics of district will have to be would candidate choice would be to calculate district that an African American chance district, using ecological The demo- inference tech- choice will win election. by niques, which candidate receives the ma- represented the district are graphics of BVAP, by jority calculated vote of Blacks.” Int.Ex. 25 at 9. percentage appears probit true that While it Georgia. Epstein’s database includes election data technique, statistical analysis is a standard Congressional, Senate and House analysis on such an no court has relied predicts He one “equal opportunity races. reapportionment plan. reviewing a data, per- number” from this and does not 1,258 separate analyses Congres- includes elec- form for the Epstein’s database sional, tions, legisla- all elections in the relevant Senate and House elections. his report, polit- he that “as a tive offices held since 1996. The database states matter behavior, voting patterns an election ical and identifies the district which held, legislative generally at the time on races are similar was the district’s BVAP election, body the race of the incumbent from one to another.” Pl.Ex. 25 at Epstein provides explanation open or the election was an seat no or whether election, winner, justification for this conclusion than special or the race of other opinion an African his that the more elections are in and whether the winner was database, the American candidate of choice. better. While footnote Epstein’s suggests statement that racial Epstein proposes that an African Ameri- voting patterns throughout Georgia are necessarily can “candidate of choice” must “sufficiently similar” to warrant the combi- be determined some element other than database, nation of election results Epstein race. defines candidate Epstein cross-examination clarified that any African American candidate choice as this footnote referred to his conclusion election, any who won white candidate perform regional that there was no need in a with a ma- who won election district analyses. Epstein Id. n. 11. testified majority jority BVAP and who received by compar- that he reached this conclusion of the African American votes cast.30 ing estimates of white crossover vote Thus, the database of elections does current districts in three 1998 any include white candidates who received statewide races. of African American majority votes BVAP, that, Epstein per- before he districts with less than 50% wheth- testified Tr., 2/4/02, p.m. probit analysis, at 44. formed his he first deter- they er won or lost. permissible that it was to treat the expert, Intervenors’ Dr. Katz contends mined Epstein’s by evaluating of a candidate of as a unified whole definition State arbitrary. voting.31 that a white crossover He uses suggests choice is He methodology ecological inference King “more natural definition of candidate of Nevertheless, thoroughly Epstein defendants cross-ex- estimates the level of African support for the white candidates Epstein American and his amined about his calculations using methodology King’s regres- a called Dr. the differences were not sta- conclusion that sion, inference, ecological analysis. or This tistically significant. methodology detail is discussed in more be- emphasizes Epstein court also that did The low. attempt rely to on the table’s calculations that, again although districts, 31. this oppos- provided to the court or Ameri- been because African 44% BVAP district that PLEx. 109. He testified ing counsel. majority of Democrat- cans would form necessary part a of his voters, was not receive the curve ic candidate would Rather, analysis. he characterized it as a probability Epstein correlates with 50% “visual explain aid” to the correlation of a BVAP. BVAP district’s with the probability that a There are very few elections in Ep- preferred candidate would be stein’s database that fall along steep Tr., 2/5/02, elected. at a.m. 15-16. The part of the S curve. two the 158 illustration is an “S curve or a probability open-seat elections Epstein does indicate curve representing the results pro- that an African American candidate of bit analysis open-seat done for the races in choice was elected in a district with less [Epstein’s] database.” Id. at 18. The than 53% BVAP. app. Pl.Ex. II (Cong. plots curve the probabilities that an Afri- 4; 89). Dist. House Dist. Of open- the 158 can American candidate of choice will be seat election data included in database, elected at a given level BVAP. This only thirty-six of represent them elections calculation is on based formula as- in districts with BVAPs of 30%-70%. Ac- sumes normal probability distribution of cording to Epstein’s database, eight elec-

the existing points. data Id. at 23. tions place took in districts with BVAPs 31%-36%, only one of which resulted in the Epstein that, testified in a district with election of a candidate of choice. After BVAP, 50% there is a probability 75% having as a served Congressionаl Repre- an African American candidate of choice sentative, Cynthia McKinney was elected will Thus, be elected. in districts with in a redrawn Congressional District that BVAPs 50%, between 44% and Epstein’s had 33% BVAP. calculations show that the probability that an African American candidate of choice Epstein’s open seat database contains n willbe elected increases from 75%, 50% ten elections in districts with BVAPs of representing steep increase in the curve. 49%-54%: five were won candidates of Tr., 2/5/02, a.m. at 29. Epstein further choice, four were not won candidates of testified “cumulative, normal choice and one was indeterminate.32 ],” “by definition,” curve[ “is steepest right There are 18 districts with BVAPs 56%- at midpoint.” its Id. at 30. 68%, all of which elected candidates of choice.

The proposed includes six districts, in which the BVAP is between Epstein admits that it prefera- would be 50.3% and 51.5%. Estimates for these dis- ble to have points data in the steepest tricts, therefore, would “steepest” be region, but political states that scientists portion of the curve. Given the absence of often do not have all the data that they tick marks on the plotted axes or data Tr., would desire. 2/4/02, p.m. at 31. points, nothing in the permits record Georgia, its Brief, Post-Trial argues court to know what corresponding proba- that the absence of such points data can Epstein bilities would assign to these dis- not poorly reflect and, fact, its case However, tricts. probabilities would the points absence of such is the result of presumably all higher 75%, than previous unconstitutional, poli- race-based 32. Specifically, one African American candi- for a found 25, district BVAP. 52.28% Pl.Ex. date BVAP, of choice was elected at 53% app. II. four candidates of choice were elected at 54% There no districts in the database be- BVAP. Candidates of choice were not elected BVAP, tween 36% 49% and no districts in BVAP, BVAP, BVAP, 49% 50.0% 53% the database between 54 and BVAP. 56% 53.5% BVAP an indeterminate result was and the benchmark under districts of ber Post- of Justice. Department of the cies greater BVAPs have plans that proposed ¶ 294. 9; PPFF at Tr.Br. in represented are His results 44.3%. than Katz, suggests Dr. expert, Intervenors’ following table. the a mar- identified have should Epstein

that his estimates. to applied to be error gin of his that Epstein states 9. 25 at

Int.Ex. the distributed, that and normally are

data is analysis probit the accuracy of

statistical curve that the fact the

demonstrated Howev- his data. 80% of over represents 18. 25 at Pl.Ex. that, to due suggests States er, United the the used explanation, without Epstein, open in the of elections large number the seat open for point opportunity” and high “equal with districts database seat According comparison. this to do BVAP, explanation an districts of percentages low a redrawn analysis, his the of prove results not to the does data the of of 80% would need the at incumbent a white with estimates district Epstein’s accuracy of an Afri- for order 56.5% BVAP curve. of the have part steepest of choice candidate American can probit that argues States' United being elected. of 50% chance sin- to even extremely sensitive is analysis cross-examination, de- In points. gle data proposed districts six are There coding of Epstein’s focused fendants 50.3% is BVAP between where senate Fourth in the victory McKinney’s Cynthia not consider does Epstein and 51.5%. an for race aas Congressional majority-mi- BVAPs reducing of effect re- was McKinney’s district open seat. majorities. bare districts nority district proposed drawn Ep- also notes expert approximately ran retained Intervenors’ she which in sin- Epstein testi- elections reflects data constituents.33 of her stein’s third at McKinney’s race Int.Ex. districts. that, taken had he gle-member fied database, point his anal- his whether explain “open-seat” does not Epstein his of out approxi- multi- proposed rise would to the applicable opportunity ysis equal of 2/5/02, at 43. Tr., a.m. in the State created points. mately two districts member plan, would cause House one race proposed Id. Thus, removal House. an districts, 15 at which point single-member his estimation there ahas members, choice 6 districts candidate two African with districts rise would winning election four with 2 districts chance 50% 3 members counts merely 46%. 44% to Epstein members. multi- proposed of seats number reap- analysis of whether Epstein’s of his com- purposes districts member consists retrogress will plans portionment majority-minority the number parison majority number comparing and benchmark under seats under benchmark districts BVAP plans. num- comparing plans, however, performed he appear, McKinney does “with stated Epstein testimony in the issue, is no there analysis, and coding this rule election, ... there’s cod- a "crucial constitutes do some what as to appropriate to record it’s think and so I a crucial that’s or not Id. ing issue.” analysis on whether 2/5/02, It Tr., at 43. a.m. coding issue.” 69 Engstrom’s 2. Ecological Use of Re- Engstrom’s report analyzes the extent

gression Analysis to Ra- Assess to which the preferences candidate of Afri- cially Voting Polarized can American and non-African American Supreme Court has relied on voters re- within Georgia State Senate dis- gression analysis to assess the severity tricts 12 and 26 have differed in recent racial bloc voting and whether existing elections, in which they presented were voting patterns prohibit would a minority with a choice between African American population from electing candidates of and non-African American candidates. Id. See, choice. e.g., Thornburg v. Gingles, Engstrom 2. reviews voting statistics 30, 52-53, 478 U.S. 106 S.Ct. 92 from five elections for State Senate seats. (1986). L.Ed.2d 25 This regression analy- include, These Senate District sis is also referred to as “ecological infer- special 1999 election, the special run- ence” methodology, political allows election, off and the 2000 general election; scientists to infer voting behavior from ag- and, in Senate District the 1996 and gregate information. Engstrom employs 1998 Democratic primary elections. Id. at types three regression analysis: ecolog- 2-3. addition, Engstrom reviews vot- inference, ical King’s ecological inference34 ing patterns in the three senate districts and homogeneous precinct analysis.35 in seven biracial statewide elections. Id. The United expert States’ report, sub- at 3. These elections include dem- mitted by Engstrom, clearly describes ra- primary ocratic for Insurance Commis- polarized cially voting patterns in Senate sioner and Labor Commissioner, the 1998 2,12 Districts and 26. U.S.Ex. 601. How- runoff Democratic primary for Labor ever, in the report submitted with the Commissioner, the general government’s election direct testimony, Engstrom for Insurance and does Labor Commissioners, not attempt predict the effect of as well polarized this as for Attorney voting General, on the ability mi- nority voters to elect Democratic primary candidates of their for Public Ser- choice under viсe redistricting Commissioner. Id. Engstrom also plans. analyzes data from three biracial eounty- Ecological 34. regression analysis provides 16%, mately an getting thus truth,” "closer to the support estimate of the for various candidates but solving aggregation inherent bias among both Tr., African problem. 2/5/02, American and non-Afri- a.m. at Aggrega- can American based statistically sig- bias problem tion is a that arises from at- nificant correlations tempting between voters' race and infer voting individual behavior voting patterns. behavior, or, One court aggregate-level from ecologi- described in other words, regression cal as a in predicting standard how statistical tech- voters of different nique “comparéis] races voted in an votes a election looking candidate at a precinct's received in an demographic election with the compo- racial characteristics and electorate, data sition of the such as voter producéis] turn-out. esti- mates of voting [minority behavior of vot- homogeneous A precinct analysis consid- and white ers] voters.” Old Cooney, Person v. ers the election results precincts that are (9th Cir.2000). 230 F.3d racially closest to homogeneous in character. “King ecological approach inference” example, For analysis generally reports technique developed in which uses percentage of the votes a candidate set all generate available data to a more accurate of candidates receive precincts within the estimate of Epstein behavior. testified which over registered 90% the voters or scholarly that a recently predicted article people receiving ballots was not African King method ecological regression im- American and within those in which over 90% proves upon by approxi- traditional methods was African American. Afri- of 90% received he 2. In Table coun- predominant the from elections

wide non- of 17.5% and votes American districts, can Senate of the each within ties lost, win- He votes.- Id. Americans African elections citywide four biracial from and 9. at Id. votes. of all .only 48.6% ning coun- these within cities largest the from analyzed Engstrom elec- Specifically, no bi-racial Id. ties. District Senate In District for Senate redis- city elections the since Savannah occurred tions elections City Macon which County County, and 2, Bibb Bibb 10. Id. at trieting. from elections and Senate District proposed the for Senate be within appears and the County its for Doughtery City, elections Albany bi-racial had District for and 162 Education, Districts the Democrat- existing House for and of Board 1-3. Tables at Attorney Id. and 12. District Senate District for primaries ic In3. at Table Id. County Commissioner. in- elections the analysis of Engstrom’s races, the Education of Board the two District in Senate Thomas volving Regina received candidates American African polar- racially of the effect 2 demonstrates American African of and 99.3% 99.5% so- of amount the minimal and voting ized Ameri- non-African of the vote, 34.2% and spe- In the 1999 crossover.” “white called for District primary Id. In can vote. African four election, were there cial candidate American the African Attorney, candi- candidates, white and two American American African of the 68.4% received Americans African of 80%36 While dates. Ameri- non-African of the vote, 16.6% and American African four of the for one voted primary Democratic Id. In vote. can non-African of 20.4% candidates, only Commission, County Bibb of the for Chair candi- four of the one voted Americans received candidate American African an runoff special 1. In Id. at Table dates. 2.7% and American votes African of 64.8% Thomas, an African Regina between votes. Id. American of non-African candi- candidate, white and American county elec- Afri- municipal of the 78.8% of date, received none Thomas In pref- did the Engstrom an vote, received analyzed she while tions American can and voters Ameri- American the non-African African of of erences 8.9% estimated coincide. election, American Thomas In the non-African Id. can vote. analysis that his of the 99.2% concludes Engstrom received to have is estimated “reveal question of the 43.6% districts and three vote American African these voting racially polarized pattern Id. vote. American non-African at 11. Id. areas.” 55% currently has the vot- attempts to simulate voter Engstrom American African 52.5% and BVAP dis- patterns at 8. Id. ing in November registration this, he considers to do order candi- In tricts. White, American an African John seven information against precinct-level and lost won date, twice African between races primary. statewide the Democratic opponent white candidates, “reaggregating” white approxi- He lost 8-9. Id. at three contours to reflect support- data Americans of African mately 65% hypo- seven Id. districts. non- 10% of him, approximately ing dis- reaggregated elections Id. thetical him. supporting Americans African ecological King’s inference testified voting pat- reports predicted Engstrom accurate most be the considered generally analysis, regression as calculated terns calculation, court will inference, homoge- method ecological Ring’s estimates. however, to these refer therefore Engstrom, analysis. precinct neous tricts, an African American candidate lost This court is properly convened as a three- only once in one reaggregated district. Id. judge pursuant court to 42 § U.S.C. 1973c at Table 4. cross-examination, On Eng- § and 28 U.S.C. 2284. parties None strom testified that there would be a contests “very the applicability of Section 5 to that, good chance” based on the this matter. patterns revealed by his reaggregation The State is a covered analysis, African American candidates jurisdiction as defined under Section 5 of would win election in the dis- reconstituted the Voting Rights 1965, Act of as amend Tr., 2/5/02, tricts. p.m. at However, 88. 42 ed. 1973c; § U.S.C. § 28 C.F.R. 51 he concludes that his reaggregation results (Appendix). Section 5 applies any to were “not a good indication that these changes ‍​‌​‌‌‌​‌​​​​‌‌​​‌‌​​‌‌​‌​​‌‌‌‌​‌‌‌‌‌​‌‌‌‌‌​‌‌‌‌​‍ in voting processes in Georgia, proposed districts provide will African and mandates that the pre- State receive American voters with a opportuni- realistic clearance prior to instituting any such ty to elect representatives choice, their changes. Georgia v. States, United 411 given that the level of crossover voting 526, 527-28, U.S. 1702, 93 S.Ct. 36 L.Ed.2d

tends to considerably be higher in these (1973). reapportionment The of seats elections than in the senate and other elec- Georgia’s Congressional delegation and tions involving local candidates.” U.S.Ex. General Assembly seats for the House and 12. Engstrom emphasizes that Af- Senate are changes in voting procedures rican American candidates consistently re- covered the Voting Rights Act and ceived less crossover local elec- require preclearance. tion than in statewide elections. Id. In an action for declaratory judg Engstrom justifies further his conclusion pursuant ment to Section plaintiff the that the local election analysis was more has the burden of proving by preponder probative than his reaggregation analysis ance of the evidence the absence of both describing second reaggregation anal- discriminatory effect and discriminatory ysis performed. he When Engstrom purpose in the reapportionment of leg its reaggregated the statewide election data islative districts. City Pleasant Grove to reflect the existing Districts, he States, v. United U.S. found that African American candidates (1987). S.Ct. 93 L.Ed.2d 866 were supported at similar levels as those reflected in Table 4 of report. Tr., his Obligations A. Court’s in Action for De- 2/5/02, p.m. at However, he noted claratory Judgment that, in the existing District Ms. The State of Georgia made the strategic Thomas barely won her runoff election in decision to an institute action in this court 1999, by votes, some 73 that, for declaratory judgment and not to seek existing Mr. White not had preelearance administrative De- succeeded 1996 or Thus, 1998. Id. he partment of Justice. The vast majority of concluded that reality did not reflect the changes in voting procedures subject prediction favorable of the reaggregation Section 5 are submitted to Department analysis. preclearance. Justice for The majority of cases that come to this represent court III. Conclusions of Law plans or changes to which Department jurisdiction court has to hear this of Justice objected. has Here, Georgia’s pursuant case 1973c; § 42 U.S.C. 28 redistricting plans have been submitted in § 1346(a)(2); U.S.C. § U.S.C. the first instance to this court for review. preclear- between difference ized judi- submitted were plans When distinc- on one based as avenues ance whether not know did review, Georgia cial and silent action affirmative between tion object would Attorney General Attorney General: by the acquiescence which did, plans which and, he if plans meth- problematic. preclearance] considered [administrative would This districts unlike § 5 is eventually identified under compliance Attorney General od of objec- preclearance judgment redistricting [declaratory the Senate implementation Sen- in that and, tionable, particular, method] conditioned is not voting laws 2,12 and 26. changes ate Districts by the At- statement an affirmative circum- unique presents This case is with- change General torney Attorney General wherein stance effect. purpose discriminatory out redistricting three to two objected § is5 with compliance contrary, the To Georgia, of the State by plans absence, for solely the measured court this to come State yet the and objection timely reason, aof whatever administrative— not judicial seeking —and Attorney General. the of part the on Unit- plans. three of all preclearance 5 Yet, Section 2411. 502, the 97 that at S.Ct. Id. argue intervenors and ed States of methods two different two presents the object clearly to to failure States’ United to point cannot declaratory The State entry of preclearance. justify does not plans acquies- apparent Attorney the retains General’s the the State because judgment the under relevant objections circumstance the cence—a because and proof of burden preclearance a administrative entry of declara- when statute preclude of intervenors judicial of grant a justify the sought that argues is The State judgment. tory —to Attorney General the Had to the object preclearance. to States the United of failure United the precleared administratively the redistricting plan Congressional plan redistricting any Congressional of court this States relieves plan House State the redistricting plan, to House respect findings with State make obligation to moot- been have may well agree. not matter does current court plans. those States, United Dallas v. City administrative ed. See to not seek chose The State of (three-judge (D.D.C.1979) to en- F.Supp. court 183 this 482 asking In preclearance. pro- not 5 did court) that (holding three Section to all declaratory judgment ter challenge to intervenors for an affirma- a forum court vide on this it imposes plans, had Attorney General the after plans voting plan the whether inquire duty to tive plan). abridg- a revised denying or approved of purpose or the effect or of race account to right vote the ing intend- Congress However, idea that the Furthermore, assumes State the color. a rubber to three-judge court this ed prepon- aby demonstrating of City the burden In simply untenable. is stamp of a declar- that such evidence States, Supreme derance the v. United Richmond warranted. is atory judgment court of this a decision reviewed Court inquiry a Section 5 engaged had as evi- v. Gressette Morris cites Plaintiff City of United States where that courts Congress intended dence decree, a consent had entered Richmond declaratory judgment for requests review to en- objected intervenors citizen but objects Attorney General only where 358, 422 U.S. decree. consent try of the U.S. 432 judgment. entry an such to (1975). 245 L.Ed.2d 45 (1977). 95 S.Ct. 2411, 53 L.Ed.2d S.Ct. Court’s this correct were If State character- Court Morris, Supreme words, jurisdiction stripped when the S.Ct. In other ap- United the new object portionment plan fails to a submitted not States to must have the effect purpose judicial preclearance, three-judge providing minority opportunity elect City court Richmond would have less candidates previous plan. choice than did the jurisdiction been inter- Id. without consider Supreme claims. venors’ While All parties in this action make liberal Court’s decision does comment on this “retrogression” use of the word in their issue, question neither did the Court arguments, shy away but any at- and, jurisdiction district court’s exercise of tempt retrogression. to define The court fact, remanded the case district can not easily so avoid this task. proceedings. court for further Id. at Section 5 does not focus on the S.Ct. discriminatory impact plan, an inquiry Similarly, reject we the State’s ar which would plan against measure the an gument that this court’s is limited review ideal, plan. See non-discriminatory Bossi challenged to those districts II, 866; er 528 U.S. 120 S.Ct. cf. States, encompass United should not States, City Lockhart v. United 460 U.S. redistricting plans entirety. in their 125, 134, S.Ct. L.Ed.2d 863 In declaratory brought a judgment action (1983) (finding that a change that did “not pursuant 5, to the court’s Section review degree the increase of discrimination necessarily extends the to entire against § ... blacks entitled 5 [was] to plan. preclear to Refusing specif the preclearance.”). objective than Rather an object ic districts to which defendants ideal, Supreme explained Court has require would nevertheless to the State retrogression that is to be measured with Moreover, plan. rework its entire Senate respect quo, to the status reflecting Con- presented authority has no legal intent gress’ existing proce- to “freeze” the that limit inquiry would the Section 5 if a proposed change dures would retro- by Attorney those challenged districts I, 477, gress. Bossier 520 U.S. at 117 Indeed, retrogressive. General as the S.Ct. 1491. Preclearance must be denied “ very declaratory of judgment structure change if a proposed ‘abridges right court, procedure, under which the and not to vote’ relative to the quo.” status Bossi- General, Attorney is with vested II, 334, 866; er 528 U.S. at 120 see S.Ct. final authority approve disapprove Hall, 874, 883-84, Holder 512 v. U.S. 114 whole, the proposed change as a argues (1994) (“The 2581, 687 S.Ct. 129 L.Ed.2d conclusively against suggestion. the State’s for comparison present by baseline is defi- nition; status.”); it is existing City Assessing Legal B. Standard for Retro- States, 125, Lockhart v. United 460 U.S. gressive Effect 132-33, 998, 103 74 L.Ed.2d S.Ct. 863 (1983) legal reviewing comparison standard (proper is between new system redistricting plans system actually prior for Section 5 in submitted effect preclearance charter, has been in a new decep adoption regardless defined Beer, In tively simple might manner. Su have Tex- required); what state law States, preme a reapportionment F.Supp. 201, Court held that as v. 203 United 785 (D.D.C.1992) plan retrogression must not “lead to a (preclearance involves a com- position change parative inquiry of racial minorities with re into whether the spect procedures retrogressive to their effective when voting exercise elec Beer, compared plan toral 425 at 96 that other- franchise.” U.S. would 74 voters....”) effect). (empha- held viously the con-

wise be force added). claim, the has sis text of a vote-dilution Court a preclearanee, which rests on held that definitions slightly While more detailed nothing “is finding non-retrogression, retrogression are found decisions that the more than a determination court, they sim- three-judge panels of this it is no more dilutive than what change ply underlying extend the rationale II, 120 replaces.” Bossier 528 U.S. must withheld from a preclearance be S.Ct. minority voting diminish that would indirectly. In Texas strength—directly or Supreme has never Court States, explained v. United the court comprehensively “retrogression,” defined preclearance rule mandates “[t]his any nor it detailed discus engaged prong denied under the ‘effects’ of Section an “effective exer sion of what constitutes minority if a voters in system places 5 new by minority cise of electoral franchise” sys- position existing a than the weaker Hall, Holder v. 512 U.S. voters. But see (D.D.C.1994); 27 F.Supp. tem.” 866 874, 895-903, 114 129 L.Ed.2d S.Ct. Reno, F.Supp. v. accord Arizona (Thomas, J., (stating that concurring) (D.D.C.1995) change (“any which tacitly the Court has selected the number place protected minority group would as its of elec of elected officials indicator position position worse than its [under strength). toral Section cases fo plan] not merit clear- the benchmark does exclusively evaluating almost cused States, ance”); New York v. United proposed change whether a would leave (D.D.C.1994) (“[i]f F.Supp. minority position voters in “worse” than no position voters is worse I, existing plan. under the See Bossier than under the new scheme it was under 487, 117 The Court 520 U.S. at S.Ct. 1491. scheme, change the old then the compliance that clearly held with Sec preclearance is entitled to under section 5, tion avoidance retrogression, does 5”). require jurisdictions improve not or strengthen voting power frequently of minorities. While courts have considered 983, Vera, 952, v. 517 116 of “majority-minority” Bush U.S. S.Ct. number dis- (1996). 1941, minority voting 135 L.Ed.2d Nor does tricts as 248 indicative require plans appar- 5 in redistricting strength, parties Section that this matter victory minority preferred ently ensure can is not an agree for Section 5 abso- Rather, it is didates. a mandate that “the lute mandate for maintenance of such dis- minority’s agreement entirely tricts. opportunity representa proper. elect This diminished, Organiza- tives of its choice not be di- It is true that in United Jewish by rectly indirectly, Williamsburgh, Carey, or actions.” tions Inc. v. State’s a Bush, 983, at plurality Supreme 517 U.S. 116 S.Ct. 1941 Court construed added); retrogression (emphasis County accord Hale v. standard to mean that States, 1206, major- F.Supp. United 496 1216 “had there been districts with black (D.D.C.1980) law, (principle nonretrogression previous ities under the and had [the requires jurisdiction] in jurisdictions that covered demon- fact covered decreased districts, minority majority strate that voters maintain “at number of it black power they pos- least as much its in modify plan electoral would have had to order system, reapportionment the existing implement sessed under” and to its carv- change enough majority in a a ing large will not result out potential advantage pre- many “retreat from the additional districts would however

75 necessary satisfy be the Beer test.” 430 some courts emphasize the need major 144, 159-60, 996, U.S. 97 S.Ct. L.Ed.2d 51 ity-minority districts to equalize voting (1977) (White, J., joined 229 Stevens, power while others have found that such Brennan, Blackmun, JJ.); accord Ket districts have the effect of isolating minori Byrne, chum v. F.2d 740 1402 n. 2 ty voters and limiting their electoral (7th Cir.1984) (defining “retrogression” as strength). “a decease the new districting plan or Thus, in an area where voting patterns other scheme the absolute num polarized race, according to distrib- representatives ber of which minority uting African American voters out of a group elect”). has a fair chance to This single district in which they majori- were a statement United Jewish Organiza ty and creating substantial minorities in a tions, however, pure plural was dicta: the larger number may districts increase ity’s holding simply was that Section 5 the voters’ ability to elect candidates of authorizes some consideration of race in choice. Hays Louisiana, See v. 936 drawing district lines. 430 U.S. 97 F.Supp. 360, (W.D.La.1996) (not- n. Moreover, S.Ct. 996. the Court’s subse ing that plan one “with its one black ma- quent cases analyzed issue jority and districts, three influence empow- creation and maintenance of majority-mi more ers black voters statewide than does” nority districts the broader context of with two black-majority districts assessing minority voting strength in a and five “bleached” districts in which mi- given jurisdiction. nority influence was reduced order to Indeed, in the cases, context of Section create the black-majority district); second the Court has observed that majority-mi Hunt, Shaw v. U.S. 947 n. cf. nority districts do not inherently increase 116 S.Ct. (1996) 135 L.Ed.2d 207 decrease voting strength, but (Stevens, J., (“Of dissenting) course, a rather can have “either effect or neither.” that unfairly ‘packs’ State African-Ameri- Quilter, Voinovich v. 146, 154-55, 507 U.S. can into limited number of dis- (1993). S.Ct. 122 L.Ed.2d 500 may tricts subject § to a challenge Breaking apаrt a majority-minority district ground it has failed to create so- and dispersing minority voters into neigh ”). called ‘influence districts’....

boring districts can have different conse such, As the Court suggested has that quences in different contexts. theOn one propriety the even hand, legality the it can minority diminish pow voters’ —and —of majority-minority depends districts on a er “fragmenting among [them] several analysis careful of the facts and circum- districts where a bloc-voting majority can use, stances routinely surrounding their particu- outvote them.” De Grandy, 512 1007, larly of at U.S. 114 nature of “society’s 2647. racial and S.Ct. On the other hand, cleavages.” such ethnic dispersal De Grandy, can 512 actually U.S. at increase 1020, 114 opportunity Therefore, electoral if it S.Ct. 2647. “pack asking eliminates ing” whereby minority whether the elimination of majority-mi- voters are a nority crammed into a small number of district —or the “safe” reduction of African deprived districts and population of an American ability to influ in such a district —is greater a Id.; ence actually number of elections. retrogressive, a court must take Hall, Holder v. 874, 898-903, 512 U.S. account of the fact that “there are commu- cf. 2581, 114 (1994) S.Ct. 129 L.Ed.2d nities in 687 which minority citizens are able to (Thomas, J., concurring) (suggesting that form coalitions with voters from other ra- 2 Section case is law inconsistent because cial and ethnic groups” and thus have less 76 affected not so is districts .the throughout single a majority within to

need “be will have that it voting bloc their racial of candidates to in order elect district avail- opportunities on the impact negative Id. choice.” American African Georgia’s able to analy- case, court’s this 5 a Section In heard. voices collective their to make retro- question of to limited sis—while care- must must we analysis, fact-intensive gression this beginning Before —is in which context that scrutinize contention fully the State’s address first occur. will changes de voting limited to proposed inquiry retrogression polarized racially of the level districts particular, reapportioned whether termining ais there to degree which op voting, “equal or the an with minority voters provide voter of a the race “between minority correlation candidates. elect portunity” votes,” voter way which plan its because that contends Georgia 30, n.&53 478 U.S. Gingles, Thornburg v. a reasonable— for black voters preserves (1986), 25 L.Ed.2d 92 of 106 S.Ct. candidates elect equal or —chance in dis- a decrease issue, on light whether sheds at districts three choice an produce will populations minority thus tricts’ Section has satisfied State “Un- effect. retrogressive impermissibly 2 test to a Section apply us to asks may districts American African packing” Quilter, 507 See, e.g., plans. consequences negative positive (“Only have if the 1149 113 at S.Ct. U.S. African strength electoral statewide effect has the scheme apportionment that vot- extent To the voters. oppor equal class the denying protected minority voters suggest ing patterns does of choice candidate its to elect tunity join forces position in a better argu 2.”). implicit § The State’s it violate to elect population segments other exist cannot retrogression is that ment candidates, a decrease preferred minority Section satisfies its where no little or may have BVAP district’s in a disagree. 2.37We De strength. voting minority effect clearly articu Supreme Court 114 S.Ct. U.S. at Grandy, 512 and Section that Section lated circumstances, dilution In such in the scheme functions separate have no retro- may age population voting Holder, 512 U.S. Act. Voting Rights However, polar- racially if effect. gressive (Section and Sec 883-84, 114 S.Ct. and its area an persists ized “structure, purpose, 2 differ tion minori- history demonstrates electoral Party Republican v. Morse application”); greatly diverge preferences ty voters’ 209-10, 116 S.Ct. U.S. Virginia, 517 voters, a decrease non-minority those of (1996) (noting L.Ed.2d lessening into a may translate BVAP of range cover broader may 5 Section strength. voting minority 2). Section than does procedures presented plans scrutinizing the highlight roles are distinct These consider we must therefore preclearance, by mandated inquiries the different by ed of its burden State met the whether Section 5. Under 2 Section Afri- Section of dispersion the demonstrating that vote’ right ‘abridges the change “[i]f population voting age American can where be denied preclearance must then, us that would Effectively, the State 37. (or Con- 2 Section plan violates proposed rejected argument adopt converse 877; II, stitution). S.Ct. at 120 Bossier See Parish the Bossier Court in Supreme by the 476-85, S.Ct. 1491. I, 117 at U.S. 520 There, Bossier the claim Court rebuffed cases. 77 relative to quo, preclearance the status is may constitute retrogression in overall mi- II, denied....” Bossier 528 at nority U.S. voting strength. comparison S.Ct. The 866. here is not Accordingly, the court’s inquiry in this to an abstract ideal such equality case will concentrate on whether proportionality, case, as in a Section 2 but State’s reapportionment plans concrete, rather to a existing plan. See will diminish African American op voters’ Hall, 883-84, U.S. 114 S.Ct. 2581 portunity to exert power electoral at the (finding no Section 2 violation because polls. While parameters of the court’s Court could not identify the “ideal” bench- investigation clear, are grounded are mark). The inquiry prescribed relativistic law, in settled this case presents a unique by the Court means that the nature of the factual context. Georgia’s State House existing plan benchmark will set the terms and State Senate reapportionment plans of the retrogression analysis. The fact were drafted to support bolster for the that a weak status quo may make preclear- Party, Democratic in part by “unpacking” easy ance does not mean a stronger predominantly African American districts. status quo should preclearance not make plan State Senate would redraw four more Supreme difficult. The Court’s dis- districts with existing of BVAPs 55.43% to tinction types between the of benchmarks 62.45% such they would have bare implicated by Section 5 and Section 2 com- majorities BVAP, ranging from 50.31 to pels this court to conclude that a plan that 50.87%. The slim majori nature these existing diminishes minority strength, but ties cause for concern when considered

which is not discriminatory, should not be conjunction levels black voter precleared. registration in the districts. In five of the “majority-minority”

That rejects the court districts posi- created State’s under reapportionment plan, tion that a Section 5 percentages violation can not exist registration black voter without a range violation of Section from 47.46 does not to 49.44%.38 mean Under question plan, benchmark whether its African registration plans are in fact levels in non-diseriminatory is ir- these same relevant districts ranges 52.48%to retrogression inquiry. See I, 64.07%. Bossier 486-88, 520 U.S. at 117 S.Ct. Evidencе that a satisfies Sec- appropriate analysis of these 2

tion preserving reasonably op- good changes focuses on they whether likely portunities for African American voters to to diminish minority voters’ ability to ef- elect candidates of may choice bear on fectively exercise their franchise. The whether there has impermissible been re- dilution, mere fact of the spreading out of trogression under Section Undoubtedly, voters, is not unlawful in the Sec- a change that has effect is likely this less context, tion at least to the extent that it to be marked retrogressive intent. does lead to a palpable decrease Yet, if existing opportunities of minority minority voting strength. See Bossier voters to exercise their franchise are II, ro- Parish 528 U.S. at S.Ct. bust, plan that leaves those (finding that an intent to dilute in violation merely voters with a “reasonable” or “fair” Section 2 is not necessarily an intent chance of electing a candidate choice retrogress of 5). in violation Section of In 38. The court does not consider Senate Dis- created "majority-minority” district. range, trict 34 in this newly the as district ais the State whether determine court must not are patterns voting if

particular, a have will not plans the that proven has voting or oth- racially polarized by marked minority on effect retrogressive of exercise the effective to barriers er strength. franchise, may dilution voters’ minority ability Strength of Voting Minority on the Measuring effect or no little candidates. preferred to elect voters those inquiry retrogression the part, In large ex- vot- fears effect plan’s to the contrary to the Accordingly, looks of number considering Act the Rights Voting by strength the ing by plaintiff, pressed the in voters that move American African adopt plans potential states to allows they in which districts. proposed and existing districts out of minorities rise gives that very the vot- data majority of a data —the census constituted formerly reap- to conduct obligation racial that divi- the State’s to provided ing population, problem, unique numeri- that point portionment presents the to healed sions have — address as must court necessarily translate the one which will and cal reductions time, the first For In power. con- matter. preliminary electoral reductions into to aver respondents prove permitted voters census that white the trast, more the race. one than for candi- of more they were their ballots that to cast unwilling list that minorities, responses and the to Thus, in addition preferred dates data sup- “black,” data includes to census decline the races that different more as “black” the candidates, identified greater the who on individuals the same port opportuni- Ironically, percent- the decreasing races. the other of impact negative identify one’s race specifically the elec- ty be on to more minority voters will age of controversy regarding behind. to a are left rise given those who has strength of toral identify a “black” voter. to best how cross-racial coali sum, then, just as In of Department on the relies States United “to pull, opportunity tion-building—the January, issued Guidance Justice political common to find haul, trade that jurisdictions notifying covered 1020, 114 U.S. at Grandy, 512 De ground,” only black consider would States United to numbers allow smaller 2647 — can S.Ct. “black to be responses and black/white its an influence, can so too great extend Yet, preclearance. of purposes for voters” drop voting, allow titheses, racial-bloc calculated redistricting office Georgia’s of retrenchment become a to numbers those indi- including as population “black” Thus, we find evidence more power. the census responded to viduals who disputed Sen polarization racial of or black as black identifying themselves persuaded districts, more we ate polit- Georgia’s While any other race. its burden to meet failed Georgia has the differ- expert suggests science ical African reductions proving methods the two inconsequential, ence districts those populations American divergent voters lead to counting districts majority-minority in other majority- the number of about conclusions African ability of its not lessened plans. minority districts effectively exercise American dispute regarding fight In right vote. collective their urges data, States United census Reapportionment Georgia’s Effect C. data registration consider black court Plans African further indication voting population redis- three each considering faced thus court is This review, the Districts. tricting plans submitted three numbers —two calculations of black registration as, BVAP as well course, of registration data —all of which present population normal movements in and out pictures different district.”); ¶ of the of proposed plan’s 6; U.S.Ex. 503 ef. 716, effect on minority voters’ U.S.Ex. 11:25-13:11; exercise of 21:9-22:3; their 25:19-26:5; 719, (Senate electoral franchise. U.S.Ex. 66:3-21 District 2 that, residents testified in their Courts have relied consistently on per- experience, black voter turnout is lower centages of BVAP to consider whether than the percentage of registered black minority voters’ ability to exercise their voters). franchise has been affected by a voting The United urges States this court to procedure change. example, For in Unit- refrain from choosing measurement, one ed Organizations, Jewish the plurality and rather to evaluate the picture, entire opinion expressed preference a voting for taking into account the different measure- age population statistics. 430 U.S. at 164 ments of eligible African American voting 23, n. White, J.). S.Ct. 996 (opinion of population. The court agrees that such voting age population Current data is pro- an approach is prudent both and in keep- bative because it indicates the electoral ing with controlling precedent. In John- potential of the minority community. son v. De Grandy, Supreme Court However, as illustrated the United rejected the idea that one “type” of num- States’ Epstein, cross-examination of vot- ber is necessarily probative more than all ing age population may not reflect other numbers: legal “The standard is not population eligible voters, may as it total population, voting age population, include college and institutional popula- voting age citizen population or registra- tions. tion, ability but to elect. The Su- hand, On the other the Court has cau- preme Court repeatedly has declined to against tioned reliance on registra- voter any elevate factual these measures to a ” data, tion may which prior reflect effects of ‘magic parameter.’ 14, 512 U.S. n. discrimination or levels of apathy (1994). S.Ct. 129 L.Ed.2d 775 community. Rome, City See 446 U.S. In light Supreme precedent, Court we n. 100 S.Ct. 64 L.Ed.2d find that BVAP may be a more appropri- (1980) (noting that courts had found ate number to consider in determining population that black figures, when “cou- whether district is properly character- pled data, with” “provides BVAP more ized as a “majority-minority” district. probative evidence ... However, than does voter its review of data, registration plan’s may predicted which perpetuate effect on black prior strength, the effects of discrimination in court will consider all the voters, information, registration of record ... including or reflect total a be- black population, among lief Negro registration population it numbers and both BVAP cannot numbers. elect candidate its choice-”). Epstein also testified that approach Such an also allows the court BVAP is a better indicator of minority to avoid the troublesome inevitably voting strength than registration numbers. question controversial of who constitutes a Tr., 2/5/02, p.m. at 49 (“Registration is “black” purposes voter for of the Voting very you volatile. If good voter Rights Act. The court is reluctant to enter drive, registration you if have different fray parties’ dispute concerning types of office, candidates running any the proper counting respons- mixed-race number of individual factors can influence es hand, to the census. On one we note analysis reaggregation testimony and lay published Attorney the General

that its bur- to meet by Engstrom, of how notice performed gave advance that Guidance purposes proposed the the that demonstrating BVAP it calculate would of den the other Act. On Rights Voting of the not have plan would redistricting all indi- to include hand, decision Georgia’s and significant Two effect. retrogressive black, themselves identified who viduals Georgia’s by presented are novel issues in combina- race or as their whether First, we report. Epstein’s on rebanee inherently race, is not other any with tion what, any, if relevance determine must fur- is decision The State’s unreasonable. of analysis for our have results Epstein’s expert Harrison’s by Dr. supported ther retrogres- plans proposed the whether de- by the unchallenged was which report, words, per- is if court the In other sive. does not court The Pl.Ex. fendants. BVAP 44.3% with a district that suaded Georgia’s States’ United that the find vot- minority electing of chance a 50% has proba- or less is more BVAP of calculation the choice, way is in what of candidate ers’ consider both Rather, will the court tive. opportu- “equal of point effect of this the considering overall calculation numbers second, redistricting plans. A retrogression? nity” of the relevant a decrease related, whether is question Plan 1. State of candidate electing a of probability the ' the Senate the evidence of evaluating more, a retro- constitutes choice, without effect, the and purpose redistricting plan’s of Beer. meaning the within effect gressive of factors range wide considers court that demonstrates report Epstein’s voting contribute that a district’s between evi- correlation is a First, reviews there the court strength. and, expert, a candidate that by the bkebhood plaintiffs and presented dence BVAP analysis pre- his statistical that district. elected in particular, choice of wib and BVAP between a correlation dicting the exis- disputed have parties None of the vot- American African of opportunity signifi- It is this correlation. of tence court The of choice. to elect candidates ers disputed. correlation of this cance argument the State’s considers also identification argues that Georgia necessi- plans were reapportionment non- American at African point which the dis- of underpopulation by the tated “equal” voters American African obligation and the State’s in question tricts candidates preferred electing of chances of principle the constitutional to abide if Afri- to determine the court permit wib reviews The court one vote. person, one vot- their will retain can racially of sparse the—albeit —evidence proposed plans. under ing strength of white degree polarized Epstein’s take court to urges this Georgia 2, 12 Districts voting in Senate crossover opportu- “point equal of a number magic lay reviews Finally, the court and 26. to each it apply 44.3% BVAP on the parties nity” testimony proffered indeed, on the elec- effect plans’ each plans and, question — their to exercise ability minority voters’ the court. currently before toral districts — franchise. electoral is untenable. approach This rejects notion The court Analysis Decreases Sen- a. Probit way any inis opportunity” equal BVAP Districts’ “point ate inquiry. the Section dispositive of Epstein’s relies State that, held repeatedly data, Supreme Court the statistical together report, while a Section suit compares the change ning for an open seat because he was voting procedures ideal, to an fair “following the open rule that seat as hav- benchmark, Section 5 actions must com- ing less percent than of [one’s] former pare proposed the plan to existing constituents,” Tr., 2/4/02, see p.m. at opportunities to elect candidates of choice. the court does not Epstein understand how Hall, 883-84, U.S. S.Ct. could fail to consider Meyer Senator von Thus, discussed, as already our analysis Bremen to be a white Thus, incumbent. focus, must not on the level of BVAP that Epstein’s “retrogression” analysis tells the will ensure a “fair” or “equal” opportunity court little or nothing about actual minori- preferred candidates, to elect but on ty voting strength in the “counted” dis- whether changes would de- By tricts. simplifying nature of the crease opportunities voters’ to retrogression inquiry has, as he Epstein elect candidates of choice. has rendered analysis his all but irrele- vant.

Epstein analyzes retrogression in two First, ways. compares he the number of We do not suggest probit that a analysis majority-minority districts under the exist- may never abe valuable tool for examining ing and proposed plans. senate pres- He retrogression, but merely that the one of- expert opinion ents no as to the statistical fered State this case entirely is significance of comparison, this and relies inadequate to that task. Presumably, pro- on no underlying analysis for compari- this analysis bit could evaluate the increases son. He merely placed the census and deceases in BVAP in each proposed data a tabular form. district and assess the statewide increase The second analysis of retrogression is or decrease in the probability that minori- based on Epstein’s “equal ty preferred opportunity” candidates will be elected un- point, and simply counts the proposed thе number of der plan. Here, however, districts with higher Epstein BVAPs than attempt 44.3% made no to address the under the existing arid proposed plans. central issue before the court: whether the comparison This problematic is proposal State’s retrogressive. is He same Epstein’s reasons “equal oppor- failed even to identify the decreases in tunity” point uninformative; it BVAP tells that would occur under pro- court nothing about the posed plan, relative increase and certainly did identify or decreases in minority voting power. corresponding in reductions electabili- ty of African American candidates of Furthermore, “equal opportunity” choice. paucity of in Ep- information comparison assumes that all seats report stein’s thus leaves us unable to use proposed Senate Districts should be ana- analysis his expected assess the change lyzed open Yet, seats. Epstein’s report in African American voting state- strength concludes point that the equal opportu- wide that will brought nity was higher much where a in- white plan. cumbent inwas office—56.5% BVAP. Sen- ator Michael Meyer Bremen, von a white While questioning the Ep- relevance man, currently occupies the seat for Sen- testimony stein’s regarding “equal op- ate District 12. Over 72% of proposed portunity point,” and noting the severe comprised would be probit shortcomings analysis identi- the benchmark district. U.S.Ex. 112. fied by the defendants, the court neverthe- Epstein’s Given own explanation of his de- less takes note Epstein’s uncontradicted cision “code” Senator McKinney as run- predicted correlations between BVAP and a candidate electing likelihood plotted correlations These choice. in his discussed curve,” and were “S

his 25, app. III. 118; Pl.Ex. U.S.Ex. testimony. and redirect cross-examination a dis- drop in testified Epstein according below, in table the reflected As in diminished a result would trict’s BVAP BVAP, only of Georgia’s calculations to American African of success likelihood dis- majority-minority existing the of one Ep- to According candidates. preferred in a decrease experience not would tricts the testimony, and his “S curve” stein’s Further- plan. proposed the under BVAP of percentages in a decrease impact of a currently has 44, which more, District it where on depending vary BVAP will majority of population majority black words, in curve, or, in other the in occurs voters, 49.21% but registered a district’s of level the actual to relation a substantial experience BVAP, also would Thus, decrease a 5.7% district. BVAP in of Dis- BVAP BVAP. in its decrease to 44.3% 50% from BVAP a district’s in majority- new proposed plan’s the re- analysis, trict Epstein’s would, to according re- theOf seat, increase. would that the likelihood in in a 25% decline sult in which districts maining twelve elected. choice would candidate over 10% decreases, reductions would in BVAP However, BVAP decline a similar follow- districts. in nine overall present if the effect the same proposed Dis- Thus, effect in Senate table shows higher. ing were BVAPs of benchmark plan BVAPs would on plan Senate where trict districts. 6.52%, majority-minority reducing BVAP BVAP decrease 63.42%, “S Epstein’s to 69.72% reduce this would that indicates

curve” electing an probability only slightly candidate. preferred African Ep- read impossible Unfortunately, it is fashion, and any accurate curve stein’s additional no provides report expert his Nevertheless, or information.39 guidance testimony Epstein’s find does court that de- suggests it insofar relevant pro- ranges within in BVAP creases districts the contested posed in (if inadequately significant may have the likeli- on impact negative quantified) bewill American voters African hood of choice. candidates their to elect able Existing Senate Underpopulation b. Districts pur- 5 for Section plan The benchmark plan, enforceable legally is the last poses 61.13%, he estimat- 95%; with district redirect, his S-curve Epstein relied On Tr., approximately 95%. probability ed that, 65.18% testify 2/5/02, a.m. at 84. BVAP, at 90% probability estimated he case Georgia’s existing which this alleged quandary insulates its from a Johnson, plan. v. Abrams 521 U.S. section challenge. See Pl.Post-Tr.Br. at 74, 97, 1925, 1939, 117 S.Ct. 138 L.Ed.2d argument 47-48. This is unavailing. (1997). Georgia argues that the demo- First, even if the State is correct graphics existing plan repre- do not some reductions in BVAP were inevit- comparison because, sent a fair basis for proposition able—a that it asserts without census, light of the 2000 the districts at attempting to prove certainly does not —it underpopulated, are all issue thus follow that Georgia compelled was to move violation of principles of one person- 2, 12, minorities out of Districts and 26 to Sanders, one vote. Wesberry See v. Indeed, extent it did. the State (1964) U.S. S.Ct. L.Ed.2d 481 actually removed some majority African (holding gross unconstitutional disparities precincts from each of these dis- population of Georgia congressional across tricts, a decision that at least casts doubt and the resulting districts dilution of vot- on its cries of inevitability.40 While the ing rights of popu- of the more residents precincts added to the districts districts). lous also significant contain African American that, The State contends where popula- communities, the overall reap- effect of the districts, tion had to be added to it was portionment is to reduce the percentages inevitable that their BVAP levels would of BVAP these districts. ¶ See, (“While e.g., decrease. PPFF at 621 As illustrated, the United States has the BVAP levels majority-minority in the there plans were alternative available that decreased, district have that was inevitable would both comported have as a with the con- result of the fact that those districts principle stitutional of generally person-one population. where short of one vote This particularly is have allowed the regard greater true with to the retention of three districts with which have numbers of African Defendants American voters in the issue.”). Georgia taken suggests disputed that this 213-23; districts. See U.S.Exs. following districts, 40. The percentages shows the table and those that were added to the registered residing of black pre- voters districts. cincts that were moved out of the Black reg. reg voters, Black Black Net % Reg. reg., of% of add. of % of net Reg. voters removed . reg. reg. votes added add. SD Deviation moved out_precincts voter added_precincts 7,835/8,626 23,225 18,866 24.37% 20.39% 14.97% 12.86% * * * 24,465 (approx. 17,533 of 42% 21.83%* 15.89% 11.8%* total added vot ers) - 10,705 17,228 9,928 17.77% 46.64% 30.28% 18.25% (approx. 91% of total added vot- ers) - 14,047 33,202 23,295 28.65% 41.80% 11.0% 20.20% * * (approx. 30,484 (approx. 19,678 of 54% 41.80% 13.09%* of 54% total added vot ers) responses present *. Plaintiff's different These are calculations. marked with an asterisk. ¶¶ 181, 326, 380; 181, 326, ¶¶ Resp. USPFF at PL’s 380. Gingles, votes.” voter which example, in Senate

313-18; For 414-19. Recogniz S.Ct. that at 54 n. however, choice, was previously of their representatives in a district population However, the accord- to district majority-minority. from district strongly vary will factors, in the including certain decreases that BVAP number of ing mere fact to a deny preclear enough to electoral allegedly not dilutive is of the districts nature Instead, 5. mechanism; under Section absence plan presence to or ance the implicated are concerns device retrogression dilutive electoral potentially other numerical appears it voters ...; registered when percentage of minority effective may dimmish changes are members who the district in Sec is free under A voting power. State of the the size minority [and] group; in a levels district reduce BVAP tion 5 to distriet[.] compli into district bring that in order to As ex- 2752. at 106 S.Ct. 478 U.S. Amendment, so Fourteenth with the ance case, above, racial present plained not limit doing it does long as in so important because critically is polarization to minority voters ability remaining Dis- in the Senate or presence its absence Bush, 517 of choice. elect candidates Cf. goes by the United States challenged tricts (suggest 982-83, at 1963 S.Ct. U.S. not or determining whether long way to no feasible alternative exception when ing Ameri- and African BVAP the decreases exists). Voting way, the Read in this are in those districts registration can voter its historical continues serve Rights to Act effects. retrogressive likely produce ex protecting minorities purpose of that, expert demonstrated franchise without own Georgia’s electoral of their ercise po- racial amounts of varying over given the Constitution’s some running afoul voting, and crossover equality. larization arching demand of choice candidates ability to elect voters’ Voting Patterns Racial c. Evidence of BVAP as districts’ to decrease likely is analysis nec- probit Epstein’s diminishes. voting “exists polarized Racially racial about information relationship essarily subsumes is a consistent where there turnout.41 and voter voting patterns way and the race of the voter between the figure whether trying to out advantage cerned great Epstein testified: "the is, here, I’m con- analysis not using probit Tr., 2/4/02, an African port. Whether American candidate of p.m. at 50. He testified necessarily choice election is that, wins deter- recollection, to his ecological re- degree mined of racial bloc by gression analysis had shown “very high African the level polarization” levels of in that election. Id. of white crossover votes received Thus, exception with the of Epstein’s However, candidate. it impossible is analysis of one election Senate District extrapolate patterns these voting Ep- 12, the only racially polarized evidence of stein’s Epstein database. As admitted on voting before the Engstrom’s court is ex- cross-examination: pert report. The report provides the re- point the whole of my analysis is not to sults of different two methodologies of an- polarization look at per ques- se. The alyzing racial voting patterns in Senate tion not whether or not blacks and 2, 12, Districts Engstrom consid- whites in general vote for different can- ered local elections to assess the “extent of didates. racial voting correlations” in the existing Tr., 2/4/02,p.m. at 44-45. Tr., 2/5/02, p.m. district. at 12. He also Although drawing any conclu- legal performed a reaggregation analysis, which sions, that, Epstein testified in order to gathered precinct-level results from seven assess there retrogres- whether has been statewide elections between an African sion voting rights people who live American candidate and a white candidate district, in a reconfigured it would be nec- predicted results for essary consider whether the district was Eng- Districts. results of “significantly different” than the rest of analysis strom’s length discussed at Tr., 2/5/02, Epstein State. a.m. at 70. the Findings court’s of Fact. that, opinion, also admitted in his a candi- running date for office in particular dis- The State would use Engstrom’s trict would want to take into account reaggregation analysis prove to case. its whether white crossover was at 25% ator argues The State reaggregation that the 52%. at 72. It Id. is inevitable that there analysis is the only record evidence that will variability” be “local in elections. attempts to estimate voting patterns in the fact, Epstein notes that his “final conclu- proposed By Engstrom’s districts. own sion ... is not that there is no difference admission, the reaggregation sug- results ... from one area to another.” Id. gest presence the of sufficient cross- white over

Epstein permit to some perform did not African Ameri- any ecological can regression analyses prevail of candidates of choice to in voting patterns in the Furthermore, redrawn did, Senate Districts 2 or Epstein Eng- districts. however, perform strom can ecological guess an at the regression reasons for analysis for an in the voting patterns election different in Senate District statewide 12. In election, However, the 1998 and Senate he found local elections. we find that that the winning white Engstrom presented candidate was not has compelling evi- choice, the American African candidate of patterns dence that racial voting in State and, fact, in very received sup- little expected Senate races can be differ to characteristics, getting

not black candidates are elected via you’re using district then esti- turnout, crossover, greater registration, looking ... mated district characteristics.... I'm appeals. campaign might districts, Those all be there at the different kinds of see who they’re important, you but if all want to elected a black candidate of choice and who Tr., 2/5/02, something do is ... estimate on the level of didn’t.” a.m. at 108-09. determining these the court would assist in statewide voting patterns from racial considered also Engstrom limits. cities’ races. but, races, no rec- again, House two State analysis discussing Engstrom’s Before compare the court permits to evidence ord elections, we in local voting patterns of Dis- the House reach of geographical the with concerns some address must first Districts. proposed Senate to tricts under- Plaintiff analysis. of this relevance any information absence scores the require do not in the record gaps These identify would report that Engstrom’s election the local disregard the court which jurisdictions of the overlap by Engstrom. analysis provided data took analyzed by Engstrom local elections portion a substantial that It is clear Districts. Senate proposed and the place Bibb Districts and benchmark 2/5/02, Tr., 15; also at see Pl.PosL-Tr.Br. districts. proposed County comprise is ultimate- the court Although p.m. at 20. Furthermore, denied the State extent the exact to determine ly unable county assertion United States’ a substan- it is clear that overlap present, ma- involve a analyzed city elections considered jurisdictions portion of tial in the benchmark the voters jority of proposed will fall within Engstrom Ultimately, issue. districts at Districts. matter lies proof in this the burden compare attempt Engstrom did not explain look to the State We the State. benchmark plan with the and to present, is not why retrogression fact, 70, and, 2/5/02, Tr., p.m. plan, vot- racially polarized prove the absence in exist- polarization racial only at looked African might diminish ing that In its Post-Trial Districts. ing Senate several districts’ light voting strength States, citing to Brief, without the United Eng- findWe BVAPs. decreased evidence, “the coun- states that any record informa- presents relevant report strom’s analyzed involve counties elections ty elections tion, that Senate and indicates all majority of overwhelming up make be marked districts will the redrawn Post-Trial U.S. at issue.” three districts Plaintiff voting. polarized high levels has searched Br. at 96. court oth- to suggest no evidence presented describing the and found record evidence erwise.44 proposed Sen- existing the overlap of Eng- critique of primary The State’s the Districts,42 overlap of as as well ate analysis of his report posits that strom’s Districts.43 proposed Senate existing and it is flawed because polarization racial city elec- However, Engstrom looked at vot- level of white crossover the Savannah, Albany, considers Macon and tions for indi- that a argues better ing.45 The State record that in the is no evidence and there District proposed Senate within the 2 is located proposed District the 42. of 65.39% existing Id. 26. population comprised of the pro- 112. of 2. U.S.Ex. 72.68% District legis- African American To extent that 44. of the comprised is12 posed Senate District of they believe candidates Id. that lators testified existing 12. District population from the under the be choice could elected racially polar- listing they suggest that plan, did not introduced United States Georgia. voting did not proposed Senate Dis- ized exist included in counties Dougherty listing indicates that This tricts. argu- two additional 45. The advances State split County is between ments, only briefly. Plain- we address which 14. U.S.Ex. Districts Engstrom's analysis is flawed argues contrast, wholly tiff that County appears Bibb seen, polarization eator of racial range is the difference in the reaggregation re- of votes received from African Georgia sults. asks the court to conclude and non-African American voters. See that the similarity “ranges” demon- ¶ that, Resp. Pl.’s (arguing USPFF strates that the level of polarization racial issue, looking “in at the actual is the same in local elections as it is in difference apparent patterns it is that elections for statewide office. The next statewide elections was indistinguish- step [sic] logic, according State, this to the is elections”). able from that the local 1-3, For to credit Table over Tables and to example, focusing instead of on Eng- conclude that African American candidates strom’s estimate that Senator Thomas re- “good winning chance[s]” of election. Tr., 2/5/02, ceived 78.8% of African American votes in p.m. at 89 (Engstrom admits suggests State that the dif- reaggregation that results show a “good ference between that number and the level chance” of African American candidate of white crossover for Thomas winning). (8.9%), is a more accurate reflection of It is fanciful to think that the court will voting patterns.

race-based defer to counsel’s “expert” alternative the-

Georgia relies on analysis “prefer- its expert ories. No testimony pre- has been ence suggest differentials” to that the re- sented to suggest that comparing “ranges” gression analyses done Engstrom on probative numbers will result evi- the local election data demonstrate example, dence. For the court does not pattern same reaggregation does his why comparison see ranges of the analysis of statewide Georgia elections. relative support levels necessarily any identifies the range of the differences in probative more than comparison support data, in Engstrom’s average local election levels of the in sup- differences posits that range corresponds this port.46 The court persuaded is not that because he failed to consider the outcomes of may despite candidates of choice win election report. racially polarized voting. the elections included in his point State belabors the that several of the argues Eng- The State of also races, report African American candidates strom's does not take into account the Engstrom level racially polarized which of African American voter finds vot- turn-out at However, primaries. ing, Democratic Plaintiff's were contention plaintiff elected. offered that, appears theory to rest on the if African testimony persuade no that would this court generally participate American voters in Dem- disprove these candidates’ victories Party primaries higher ocratic at a rate than racially polarized existence of voting. Rath- voters, they other are able to influence er, Engstrom testified that the fact that an primaries, outcome of the the Demo- gener- African American candidate won in the likely cratic gen- candidate was to win in the change any al election would opin- not of his However, eral election. there is no reliable respect racially po- ions with to the levels of turn-out, concerning record evidence voter Tr., 2/5/02, voting. p.m. larized at 86-87. primaries general or at elections. explained He further that the victories were typically majority-minority jurisdictions. in By way only, pro- illustration the court of Contrary suggestion

Id. to the of our dissent- following average vides the calculations of ing colleague, way the court in no considers support differences in for African American by electoral victories African American candi- preferred candidates in those elections listed aberrations; rather, dates to be the can court by Engstrom in Senate Districts and and that, given specific demograph- conclude County, appears primari- in Bibb which to be voting patterns, ics and ly, totally, African American if not within District 12. Average percentages difference in of Support African American and Non-African American for African American Preferred Candidates 2_District District 26 racially po- not be will districts contested ranges the comparing method of

the court can Consequently, the coun- by plaintiffs larized. suggested differentials of the evidence open preponderance to the by a probative. While find is at all sel and in levels of in BVAP difference reductions planned that the the that contention support regis- voter American African American of African number white in the degree of racial 12 and probative of Districts may be voters in Senate tered way that it sees no the court vot- polarization, Afriсan American diminish 26 will not informa- interpret this may competently Once strength these ing .districts. to expert evidence absence of tion in the case may it well be we note that again, effect. this in African American any that decrease 2, 12 Districts power in Senate electoral with the State’s problems light of by gains other 26 will be offset inability to its statistical evidence own present districts, failed to plaintiff has but presented doubt on that significant cast any such evidence. States, compelled we the United polari- of racial the evidence that conclude Testimony Lay d. retrogres- of suggests likelihood zation Indeed, despite importance sion. jurisdiction history of The electoral inquiry, to the Section information such on the effects light may shed considerable no the court with provided plaintiff has minority will have on reapportionment that information re- comprehensive competent, part of that electoral voting strength. One voting or levels of crossover garding white patterns in racial history is found districts across in individual polarization voters to cast ballots willingness of and the time, the United At the same the State. Lay races. wit- of different for candidates evidence that produced credible has States race the use of testimony regarding ness racially highly suggests the existence is also relevant. politics elections districts. polarized voting types of presented with two The court is above, African American an emphasized As cli- the electoral testimony lay regarding jurisdic- in a ability to succeed candidate’s districts and disputed mate of polariza- levels of depend on the tion will (1) testimony general: legisla- State Eng- voting. and white tion crossover drafting in the and others involved tors that Senate testimony suggests strom’s (2) reapportionment plan; and testimo- districts will be races providing and citizens ny by legislators voting. This by racially polarized marked voting strength. opinions about utility Ep- undermines evidence Support Legislators’ American African analy- probit analysis because stein’s Redistricting Plan variations levels fails to account for sis heavily on the unanimi- The State relies plaintiff And polarization. racial support legislators’ African ty of persuade no other presented evidence argue reapportionment races in the voting in future us County Statewide Statewide Bibb Senate 12 Statewide elections elections elections elections elections elections _(primaries) 28.43%55.72%40.2% 63.58%45.99%63.05% Engstrom's Ecological Tables King Inference numbers contained on the differentials are based The 1-3. Eng- contained in King logical Inference numbers Eco- based on the differentials are The plan retrogressive imposing does not have a an electoral scheme which splinters “indisput- geographically effect. Plaintiff characterizes as concentrated populace proposition racially polar- that African Ameri- within a able” parish, ized thus representatives minimizing can elected are in the best the black citizenry’s participation. electoral position judge “as a matter of fact” reapportionment plans whether en- Treen, Major v. F.Supp. minority hance diminish (E.D.La.1983). case, In this the State has ¶ court, strength. PPFF 618. The how- presented authority no suggesting why the ever, notes that the United States has support court should consider the of Afri- presented extensive evidence African can legislators American as evidence that misgivings Senators’ about the the actual of the Senate redistricting effect Nevertheless, plan. Afri- two plan will not be to decrease vot- legislators against can American voted opportunities ers’ to elect candidates of plan. The court will not look behind those choice. We believe that the legislators’ question inherently political votes to such is, end, support probative far more decisions. A for legislation vote is almost of a lack of retrogressive purpose than of sort, always compromise of some moti- an absence of retrogressive effect. complex vated intersection of self- Voting Patterns in Senate Districts pressures. interest and external A court and 26 forces, unpack that tries to these and as- replete lay record is with testimony them, probative sign weight to treads a regarding predictions individuals’ as to path. Accordingly, treacherous we are whether the redrawn permit districts will rely testimony loathe to regarding the minority voters to elect candidates of trade-offs, legislative post- nature of or on choice. Several incumbent Senators testi- expressions part hoc of doubt on the they plan fied that believed the gave them legislators who nevertheless voted for the and others a “fair” or “reasonable” chance plan. Certainly, contested as it relates to See, victory in the redrawn districts. effect, plan’s possible retrogressive e.g., dep. dep. Brown at Walker at 12. this is dubious evidence indeed. However, expressed Fort Senator con- said,

That undoubtedly it is true plans cerns that the might retrogres- support among minority legislators sive. Int.Ex. 16 at 3. reapportionment plan for a can sometimes The United States offered the testi- assessing be relevant in legality of that mony of eleven witnesses from Chatham Voting Rights under the Act: County respect Sen- protection existing relationships ate including District two State Sena- among incumbents and their tors, constitu- four African American Commission- ents, accruing and the benefits to the ers, and three members of the Executive state from the seniority delegation its Committee of the Savannah Branch of the may Congress, have achieved in They they NAACP. testified that believe pragmatic considerations which often Proposed the boundaries of Senate District figure prominently drawing will opportunity reduce the that African congressional have, districts. These consider- American voters under the bench- talismanic, however, ations are not and mark to elect candidates ¶ may protect not serve to incumbents of choice. USPFF at 187.

strom’s Tables 1-3.

90 sta- member, a television worked at who Jackson, of the Savannah President

Dr. routinely relied tion, the staff testified that NAACP, to the a letter sent Branch of the 2001, predict 6, to voting patterns upon polarized August Governor that had not precincts to the results from opposition election expressing his ¶ 504, 6 they that “would suggesting reported their returns. U.S.Ex. changes, and (Johnson). votes and testified to the Black witness only to dilute Another serve Legislators Black voters appeals to eventually delete of racial white the use 504, ¶23; U.S.Ex. and mayoral U.S.Ex. election this area.” the 1995 Savannah ¶¶ (NAACP member 504, 10, 24 executive municipal election. the 1999 Savannah (Aider- ¶¶ districts would testifying 211; that 210, 712 U.S.Ex. USPFF strength); U.S.Ex. minority voting Jaсkson) opponent weaken (stating that white man ¶ ex- (same). city aldermen 509, 10 “Farrakan Jackson labeled Alderman opportu- concerns that pressed similar supporter”). to voters elect American

nities for African 12, the District respect to Senate With will dimmish. choice candidates testimony of five offers the United States ¶ 505, ¶5; 501, 5; U.S.Ex. U.S.Ex. declarants, one former African American ¶ 502, 16. U.S.Ex. Commissioners, a City two current from Senate Two witnesses House, State Representative to the former reg- African American voter that testified Doughtery president former the white voter is lower than istration rate witnesses, Two of the County NAACP. Count, lead- rate Chatham registration Sherrod, and Mr. Charles White Mr. John vot- registered majority to a of white ing 12 seat. ran for the Senate District American ers, African despite majority ¶ 716, 6; U.S.Ex. population. U.S.Ex. Williams and White City Commissioners (Jack- 21:9-22:3; 11:25-13:11; 25:19-26:5 felt the flier distributed they testified (Riv- 719, NAACP); son, 66:3-21 U.S.Ex. Taylor, Mark opponent, by White’s white Bd.Comm.). ers, Fur- County Chatham Party course of the 1996 Democratic witnesses, including thermore, seven of attempt encourage to an primary, was members, County NAACP Executive Amer- vote African against white voters to Commissioners, Aldermen and Senator ¶ 12. U.S.Ex. ican candidates. Thomas, voter turnout white testified Williams, who was compared to flier White higher than African generally city issues of contracts outspoken on County.47 in Chatham voter turnout businesses, action affirmative black-owned housing, city government affordable that, experi- in her

91 that it statement would “more lay witness evidence Jones’ The United States’ paint a to elect a candidate of choice District 26 does difficult” from Senate charged political in racially minority a to a vot- picture points retrogression of such lay environment, suggested by as is ing strength. 2 Districts and 12.

testimony about Senate Legislative Influence incumbent Brown is the Robert Senator due to an required argu- A final note is opposition reelected without and has been by Georgia that rests on ment advanced in times, opposition once without three expectation proposed that the districts witnesses be- election. Most general Georgia candidates. will elect Democratic Brown incumbent Senator lieved that asserts that African American voters proposed under the would be re-elected Party tend to vote for Democratic however, especially fear- they were plans; in excess of 90% of the time. candidates to elect a candi- it will be difficult ful that appears unsup- to be While this assertion Dis- proposed Senate date of choice evidence, ported by empirical Epstein pro- did not decide to if Brown trict 26 Senator ¶ footnote, Pl.Ex. this number 7 vides re-election. U.S.Ex. run for Ed.); plaintiffs lay at 17 n. and several of (Barnes, County Bd. of U.S.Ex. Bibb ¶ (Hart, Af- 8; support proposition 36:15-37:3 witnesses U.S.Ex. Comm.). from Sen- County overwhelmingly The witnesses rican Americans vote for However, expressed concerns about ate District 26 candidates. it does Democratic in the district’s the reductions anything good that is not follow that people tend to population and testified that Party good for African the Democratic See, e.g., racial lines. U.S.Ex. along vote least, the con- American voters —at within ¶ (Abrams, 6; 723, 53:11-16 U.S.Ex. inquiry. text of this court’s Section Ed.); Bd. of U.S.Ex. County Bibb equate court to African Georgia asks the (Bivens, 28:11-29:10, County 82:22-85:5 strength electoral American voters’ Comm.). However, these witnesses did Party the Democratic at the the success of appeals racial type not describe the in this case is abso- polls. “The evidence District polarized voting evident minority voting lutely uncontradicted that 2 and 12. the Democratic strength is enhanced argues The that admissions State and white joined by both black strategy, it States’ declarants United Democrats, major- Democratic to maintain possible to elect candidates of would be ¶ PPFF at 618. ity the Senate.” proposed districts are evi- choice crafted reapportionment plan has been occur. retrogression that no will dence continued Demo- predominantly to ensure ' Jones example, For Alderman Clifton Senate, and the State cratic control of that, under “although more difficult states “political perfor- plaintiff explains that plan for blacks the new Americans mance” statistics for African to elect the opportunity have a fair still redrawing of the districts. influenced the dep. their choice.” C. Jones candidate of However, analysis of racial Engstrom’s However, reliance on at the State’s that white demonstrates voting patterns misplaced is its is as these statements regularly American voters and African point.” “equal opportunity reliance on the candidates prefer different Democratic not ask if analysis does retrogression Party primaries. U.S.Ex. Democratic “fair” or whether a district is Furthermore, record Tables 1-3. minority preferred possibility exists that Indeed, many of the demonstrates that elected. candidate would be legislature. I think that’s cur- trolled state protect were drawn to senate districts a must. important. I think this is Democratic incumbents. See rent white (Senator 703, 49:7-16, 80:4-81:20 U.S.Ex. (Congressman Lewis Pl.Ex. 18-21 white that he considered Brown testified Test.). link in the Democrats to be the weakest pointed authority to no The State has that he excluded redistricting process, and proposition politi- that a supports precincts from majority African American success, accompa- party’s cal overall neighboring in order to assist his district power minority leg- for nying positions of incumbents). Democratic

white islators, assessing in should be considered of their minority voters’ effective exercise urges Plaintiff the court to consider Indeed, Whitcomb v. Chavis franchise. in influence” as a factor assess- “legislative There, otherwise. the Court suggests minority voting ing of the effectiveness “racial specifically distinguished between leg- African strength. All of the “political dilution” and defeat at the vote in currently serving islators 124, 153, 1858, 29 polls.” 403 91 S.Ct. U.S. Assembly are Democrats. Sena- General (1971). L.Ed.2d 363 The Court concluded that, Republicans tor Brown testified were evidence demonstrated a lack of that the Senate, in majority Demo- community. voting racial Id. bias African Americans would lose sever- cratic suggested in the record that Afri- Nothing See, e.g., positions al as committee chairs. prevented can American citizens were Congressman Pl.Ex. 20 at 23-24. John vote, from were excluded registering Lewis testified to similar conclusions: political parties, or were overlooked in happen I to believe that it is the best slating process. Id. at the candidate voters, of African American interest 149-50, 91 S.Ct. 1858. Had the Demo- concerned, far I’m to all voters and as as elections, more con- crats won Court citizens, all to have a continued Demo- cluded, community Africаn American legislature Georgia. cratic-controlled complaints would have had no about its all, I great majority, First of representation. Id. Because most of the legislators present- think all of the black were elections under consideration won elected, ly and most the would-be however, Republicans, “the failure of the candidates, are Democrats or will community] legislative have [black seats Assembly. be Democrats the General proportion population^ emerges to its So, officials, if black elected black state losing more as a function of elections than legislators, going any to have sense against poor Negroes. of built-in bias control, any power, sense of as chairs voting power ghetto may residents of full committees or chairs of subcom- ..., been ‘cancelled out’ but this seems mittees, being it will come with euphemism political more a defeat at majority. minority, And when in the polls.” Id. at 91 S.Ct. 1858. they They all of that. lose lose League United Latin American chairmanship possibility maintaining Clements, the Fifth con- Citizens v. Circuit full of a committee or subcommittee. a claim that a sidered Section violation great majority plaintiffs argued of the African Amer- was shown where Georgia, ‍​‌​‌‌‌​‌​​​​‌‌​​‌‌​​‌‌​‌​​‌‌‌‌​‌‌‌‌‌​‌‌‌‌‌​‌‌‌‌​‍highly partisan voting ican voters the State of was indicative of percent racially polarized voting or more tend to the Demo- because the ma- vote So, way. jority Republicans cratic it’s in our best interest were white. 999 (5th Cir.1993). concluding for us to maintain a Democratic-con- F.2d 831 While

c© CO searching in a prox- engaged We have review parties did not serve political race, of the record for evidence that would facili- agreed the court id. at ies for competent comparison tate a of the bench- summarily dismiss should not that “courts proposed plan, mark and the racially in cases where dilution claims vote consequences and their for the patterns correspond with divergent voting strength Georgia’s of African American ‘political defeats’ not affiliation as partisan population. conclude that the State § We under 2.” Id. 860-61. cognizable proof. its of

has not met burden the converse is believe that We expert presented by plaintiff the testimony voting pro change a also true. Where pre- woefully inadequate. Epstein was party sup political will favor a cedures predict report sented a that was crafted to Americans, need by African courts ported “point equal opportunity” a of that has that African American voters’ not conclude retrogression inquiry little relevance to the corresponds to that success. strength by mandated 5. Plaintiffs retro- Section Circuit, Rather, Fifth courts as did the analysis simple consisted of a gression inquiry their obligation an to base majority-mi- comparison of the number of searching practical evaluation “upon a districts, nority and the number of dis- ” present reality.’ Id. at 860. ‘past “equal” electing with chances of an tricts in this 5 matter re inquiry Section preferred American candidate. African opportunities quires us to scrutinize expert only marginally Defendants’ was in the redrawn districts helpful. provided a re- Engstrom more franchise. The to exercise their electoral conclusions with port presented that no rejects the that a emphatically notion court retrogression respect to the existence incumbents plan protects that Democratic plan. Finally, intervenors’ ex- the Senate majority necessarily is and a Democratic offered evidence at all. pert no substantive respect with retrogress that does not short, by presented In the evidence voting strength. to African voluminous, may relied parties, while Democratic success the Whatever limited conclusions. upon only for several of the Senate Party may enjoy as result Districts with ma- The number Senate im redistricting plan does not and cannot would, jorities according Geor- of BVAP racially retrogressive ef plan’s munize the calculations, from twelve to gia’s increase Voting 5 attack. The fects from Section thirteen; Attorney according to the Gener- safeguard Act was not enacted to Rights data, census interpretation al’s any particular the electoral fortunes twelve to decrease from number would party. political pro- According Georgia, eleven. 2, 12 and 26 would

posed Senate Districts e. Conclusion 50.81%,50.22% and 50.39% have BVAPs of calcu- respectively; Attorney General’s court to enter a declarato- asking In this less, slightly with lations of BVAPs reapportion- ry that its Senate judgment falling below 50% Senate the effect of plan does not have ment Senate 49.81% BVAP. Under opportunities to worsening minority voters’ Districts plan, the number voting rights, effectively exercise their will registered voters majorities of black competent evidence present must State from 11 to 8 districts. decrease engage in the permit that the court to will that, predicts as BVAP expert Voting Rights Plaintiffs by mandated analysis decreases, probability that a district Act. decrease suggest no evidence to this an African American candidate of choice necessary. percent- will be elected diminishes. The was inevitable or existing in 11 of 12 of the ages of BVAP conclusions, court fight In of these by majority-minority districts will decrease finds that the State has failed to demon- the redrawn districts. 3.42%-26.28% preponderance of the evidence strate Districts 12 and it will de- reapportionment plan for the 10.27%, 4.77% and 11.65% re- crease retrogressive not have a State Senate will spectively. simply persuaded, effect. The court analysis regional An of local and elec- it, the basis of the evidence before racially presence tions demonstrates the minority voting strength signifi- will not be polarized voting in the benchmark Senate cantly by proposed diminished redis- in Districts and several of the counties plan proposes The decrease tricting. to in cities included the benchmark Senate existing majority-minority in the BVAPs portions of these Districts. Substantial they such that would constitute districts jurisdictions proposed fall within the Sen- majorities, slightly bare less than Lay testimony ate Districts. witness also majorities. Georgia’s It was burden to presence racially polarized suggests the produce prove some evidence to that these voting, especially in Senate Districts changes retrogressive. not be would is, however, conflicting lay tes- There timony probability that the regarding produced The State has no evi permit Afri- redrawn Senate Districts will demograph dence to demonstrate that the can American of choice to win candidates proposed ics of the Senate Districts coun election. in any teract reduction BVAP. It has not attemрted

Reaggregated results from statewide to show the number of white elections show that African American can- voters who cross over to vote for African may garner didates sufficient white cross- in American candidates of choice the dis in over votes to win election puted might districts and how that affect However, Districts 12 and 26. the effective exercise of voters’ supports also a finding record presented Nor franchise. State African American candidates of choice run- regarding potential in mi gains evidence ning unlikely for seats are to State Senate nority strength in voting Senate Districts receive the levels of white crossover same other than Districts and 26. There may voting as occur statewide elections. are, doubt, ways, without numerous other given racially polar the limited evidence of Finally, the decline in BVAPs ized State Senate and local elec 2, 12 required by Districts and 26 was not tions, could have met its bur obligation the State’s constitutional to Yet, proof den of in this case. the court is comply princi- draw districts that with the reviewing present limited to the evidence ple person-one of one vote. Under the parties, compelled ed and is hold to reapportionment plan, predominantly Afri- not met State has its burden. precincts can American are removed from Accordingly, we are unable conclude underpopulated Senate Districts reapportionment plan that the Senate will Although plan proposed and 26. retrogressive have a effect on the vot population add African American to these districts, ing strength Georgia’s African the total result is a decrease in American the districts’ BVAP. The State has offered electorate. by preponderance establishes Congressional Plan the record States

2. United re- Congressional that the of the evidence three African Georgia has Although districting plan does not result retro- in the United representatives position of African Ameri- gression currently only there is Congress, States the effective exercise respect cans with dis- majority-minority Congressional one franchise. of the electoral trict, Congressional District. the Fifth in- redistricting plan would Congressional majority-minority clude two House Plan 3. State argues that Plaintiff districts. data, the According to the 2000 census Afri- districts where creates additional also plan contains 40 benchmark State House opportunity have the can Americans population districts with a total black over election, increasing BVAPs several 50%, 37 districts with total BVAP over districts. 50%, and 38 districts with total black vot- Congression- Intervenors assert that ing registration over 50%. redistricting plan violates Section al First, Congressional that the they argue proposed plan would have 38 or majority-minority two plan does not create majorities in districts of BVAP. seats Rather, African American when districts. existing some of the House districts While in is counted accordance population in un- experience would decreases BVAP Guidance, only Attorney General’s proposed plan, der the there is no evidence majority-minority District remains a Fifth racially polarized voting the court of before However, has a the Fifth District district. any might suggest in House Districts that persuad- court is not majority BVAP. The retroges- these decreases will have a that African Ameri- the reduction ed that effect. sive neces- population in the Fifth District can sarily retrogression. constitutes that the House Plan argue Intervenors

Second, contend that infirm it contains multi-member intervenors because retrogresses proposed Congressional plan that They suggest courts districts. Fifth Dis- of the reduction because multi-member dis- disfavored generally BVAP, racially purported and due to trict’s po- such districts have tricts because polarized voting in one 1998 election minority “submerge” to substantial tential and one 2001 County Fulton Commission minori- making it difficult for populations, City president. Atlanta Council runoff for Howev- candidates of choice. ties to select however, is, in the rec- There no evidence er, districts of the multi-member support to these assertions. ord plan, majority-minority six ranging from 54.14% districts with BVAPs

Finally, suggest intervenors dis- to The other multi-member 65.18%. plans proves presence of alternative BVAP, relatively low levels of tricts have retrogression. The fact the existence of Plaintiff from 0.58% to 33.52%. ranging plans result that some of intervenors’ will that African American voters asserts in certain higher percentages of BVAP majority- strength in the voting maintain others) (and percentages lower districts districts, they minority multi-member Georgia’s proposed does not establish that candi- opportunity to elect will have plan retrogresses. Interve- Congressional Intervenors to each of the seats. dates evidence, best, shows that nors’ arguments no or present evidence adopted a Assembly could have General multi- that the convince the court plan. The evidence would Republican more jurisdiction’s prima showing facie minority will diminish member districts proposed voting change does not have a voting strength. purpose pre- in order for retrogressive House redis State denied.” Id. at clearance to be create additional tricting plan would two S.Ct. 866. majority-minority districts. The seats making has the burden of persuaded court is that the House redis prima showing plans that the do not have a tricting plan retrogressive will not facie purpose. have a Id. The retrogressive minority voting strength. effect on blatantly partisan nature of the redistrict- Reapportionment Purpose D. process goals and the of the authors of ing Plans redistricting plans do not violate Sec- help An tion 5. intent to the Democratic argues that all three redis- The State purpose, an Party retrogressive is not a tricting plans were drawn with the intent intent decrease Democratic to maintain and bolster the strength. Id. control of the State House and Senate. goals, order to reach these Democratic Here, plans undoubtedly the State’s legislators sought “unpack” majority spread Georgia’s African American out districts, and to increase Democratic Plaintiff, however, contends, voters. voting strength throughout the State. The disputed, no one has that this was done not testimony legislators Democratic State purposefully minority voting decrease reflects a belief that African American strength, but instead to increase the elec- votes, generally heavily favor Demo- which opportunities Georgia’s toral Democratic crats, high are “wasted” in districts with Party. Those individuals involved in the African percentages of American voters. process redistricting length testified at re- ¶ (“In- See, e.g., Resp. Pis.’ to USPFF garding their intent to bolster the Demo- creasing the District 12 BVAP majority in Party’s Georgia cratic Gen- and diminish would waste See, Assembly. e.g., eral Pl.Ex. 20 at 23- impact in their other districts and the re- legal infirmity see no this ef- We apportionment plan of the as a State partisan fort. progression Just as is no

whole.”); (Meggers). Pl.Ex. 22 at 21 guarantor compliance, neither Section plans prevent reflect an effort to also Moreover, proscribed. is it this court’s placed from being Democratic incumbents previous analysis makes clear that the di- in the same districts. minority lution of voting blocs does not II, Here, always retrogression. constitute Supreme Bossier Court “ end, prohibit preclear simply § held that there is no evidence that 5 does not plaintiff redistricting plan opportu- ance of a enacted with a intended to diminish the discriminatory pur but nonretrogressive nities of voters to elect candidates such, pose.” 528 U.S. 120 S.Ct. of their choice. As the court con- 145 L.Ed.2d The Court that its Georgia’s held cludes that neither United States “longstanding interpretation” Congressional plan of the effect redistricting nor its prong plan as limited to a concern with retro State House was enacted with an applied “purpose” to gression prong impermissible retrogressive purpose. “only which must cover retro Section While the same record evidence would gressive support finding dilution.” Id. at 120 S.Ct. a that the State Senate action, therefore, plan 866. In a Section 5 a retrogressive pur- does not have government pose, “need refute the covered we need not reach this issue as we has demonstrated a The State plan retrogres- that the Senate hold by preponderance a of the evidence that sive effect. redistricting House does not State purpose denying or effect of have the CONCLUSION IV. vote abridging right to on account of having considered the en- Accordingly, Accordingly, declaratory a race or color. the relevant statuto- record herein and tire is that Act No. judgment GRANTED law, hereby it is ry and case 2EX23 does not violate Section 5 of the that Civil ORDERED Voting Rights Act. par- motion for leave to Liberties Union’s Judgment accompanies An appropriate DE- is [108-1] amicus curiae

ticipate as Opinion and Order. this NIED; it and is HaRry Judge, con- T. Circuit Edwards, plaintiffs that ORDERED FURTHER curving, with whom SullivaN, of inter- grant the court’s motion to vacate I Judge, fully Judge concurs: concur DENIED; and it is vention [92-1] comprehensive opinion for the Sullivan’s judgment portions of ing plan would to intervene [144]. holds consider are DENIED. tunities to exercise their electoral have the plaintiffs request dants’ race or color. evidence that the State demonstrated effect on African abridging This FURTHER ORDERED For the proceedings the court DENIES polls. Accordingly, and intervenors’ court is divested Michael purpose that Act No. the State of Epstein’s foregoing right ^[145] not have a King’s motion for a American voters’ or effect of to enter a preponderance of the to vote on and renewed motion testimony reasons, motions to strike Senate redistrict- Georgia has not 1EX6 jurisdiction retrogressive that defen- declaratory denying account of the court [149][150] does power at oppor- stay not or to wrong court. write minorities ly done is into the colleague, for this serves an covered this court has been choice. This argues made of the law. What the dissent durally and considering that § First, 5 case. weigh the various forms of evidence been it clear that on the law and arguments I Supreme jurisdiction argument, which is the law: Our § presented “equal whose The dissent cites no to elect candidates import inquiry. § substantively is not an accurate statement separately only 5 is satisfied whenever no or fair made Court § opinion charged adopts plan § to this court. such 5 and misguided 2’s focus on dissenting colleague This is has opportunity” authority § 2 are distinct our is, not has effective- not the adjudicate consistently to I dissenting surprising authority of their respond equality believe, to proce exists. provi that pre- how way a Georgia has demonstrated The State of Parish Reno v. Bossier School sions. See that Bd., 471, 477, of the evidence by preponderance a 117 187 520 U.S. S.Ct. ”). (1997) (“Bossier redistricting plan does Congressional I Section 730 L.Ed.2d denying purpose minority every or effect of voters in state protects not have the 2 plans deny account of that them right to vote on electoral abridging against or declaratory voting opportunities. fair See Accordingly, equal color. or race or 146, 155, Quilter, 507 U.S. that Act No. v. is Voinovich judgment GRANTED (1993). 1149, 122 L.Ed.2d 500 5 of the 113 S.Ct. 2EX11 not violate Section does contrast, protects minority Section Act. Voting Rights existing plan protected group affords the in certain southern states from voters equal than chance of plans significantly better changes in electoral that have electing that same slate of candidates. Ac- their purpose diminishing or effect of vot- things being equal, all other cordingly, rights quo relative to the status that is ing that a safe district into one state converts changed. to be See Reno v. Bos- only where African Americans have a “fair Bd., 320, 334, 528 U.S. sier Parish School hard-pressed to (2000) opportunity” would be 145 L.Ed.2d 845 S.Ct. § preclear plan analysis under the its (“Bossier ”). II by Supreme For it described Court. It is therefore clear that the standard simply plan could not be said of such a against plan which a must be measured plan] it “is no more dilutive than it [the preclearance always §a the ex- case II, replaces.” Bossier 528 U.S. at illusive, isting plan. The standards of S.Ct. equality proportionality are not sum, then, Supreme neither § guideposts in a 5 case. The electoral court, any Court nor other held—or opportunities under the bench- available § preclearance even hinted —that under plan inquiry mark frame the determin- protects granted plan must be to a corresponding opportuni- ing whether opportunities equal electoral § by plan satisfy afforded ties the new voters, materially it though even reduces protective A quo status of African Ameri- opportunities those available under the ex- voting strength can cannot be weakened electoral This is isting hardly scheme. merely quo because that status exceeds remarkable, concept as the entire of retro- requirements. minimum That law’s gression against militates such a result. preclearance can be denied when a dissenting colleague simply Our blinks le- plan existing opportunities retrenches on gal reality insisting otherwise. hardly suggests, as the dissent would have it, that substantive retrenchment can ^{[*] [*]} # ^{[*] [*] [*]} Instead, simply ignored. certain cases legal This error infects the whole of the one-way imposed § ratchet 5 means analysis. Discounting dissent’s the need tangible gains made African Amer- retrogression for a inquiry serious leads icans need not be surrendered testimony the dissent to overvalue the merеly sought because the State has legislators the African American who gains (per- undo those with a that is spoke plan. favor of the haps) independently unlawful under dissenting opinion suggests that this probative evidence is more than what it § 2. expert testimony describes as the “flawed To so hold would be to undermine the conflicting lay testimony” pre- witness Supreme Court’s consistent efforts to con- by plan’s opponents. ap- sented the This § § 2 and strue 5 “to combat different proach has several serious flaws. and, accordingly, impose very evils to dif- I, upon place, ferent duties the States.” represents inap- Bossier the first it an propriate attempt 520 U.S. 117 S.Ct. 1491. the When to reframe the case with retrogression idea seriously, politicians is taken as the testimony of at the center of do, quite the dissent refuses to it is obvi- inquiry. highly proble- the court’s This is matic, a proposed plan ous that backslides from parties because all of the to this that, an existing merely litigation if it affords the in agreed light of the (and protected class an equal opportunity apparently to a State’s deliberate strate- gic) elect a fixed number of candidates and the significantly decision to decrease the politicians’ the testi- turning to Before in several Senate voting populations I weighed, how it first mony and should concern- districts, the statistical evidence assessment of the im- reject in the dissent’s issue principal is the ing polarization clarity and the statistical portance agreed parties generally case. The the only The in record before us. evidence the we that, polarization, to assess order was presented by Georgia evidence African such degree to which the must consider analysis. For this Epstein’s probit Dr. on white rely had to have American relevant, all it was crucial to be at analysis candi- preferred to elect votes crossover that it was to demonstrate for the State framed both issue As the was dates. of racial assessment States, based on realistic these data and the United Georgia Otherwise, Epstein’s Dr. tes- polarization. fac- the determinative highlighted were utility in virtually no deter- plan timony is of the Senate whether assessing tor the numerical decreases mining whether certain- parties precleared. The should be actually will disputed affect the districts the inconsistent suggest that ly did not Yet, strength. voting African American politicians African American testimony of assumption for the State’s the entire basis the statistical overcome somehow could in these areas data polarization of low polarization. on data races, which the defen- from statewide Indeed, evidentiary case is the State’s rejected as expert, Engstrom, Dr. dant’s racially polar- that on the assertion built degree likely on the bearing little having state, uniform across ized local Senate elections. crossover in of white Epstein’s Dr. utility The uniformly low. put forward as analysis, which polarized that Engstrom Dr. testified evidence, rises and falls centerpiece of its persist in local elections voting patterns proposition. of this validity on challenged districts the Senate view its own this and ignores offers dissent that in indicated His data States. United should plaintiffs case on how Dis- races nonstatewide words, the dissent In other argued. been 2, 12, polarization racial was tricts it the State what attempted to do for has persis- crossover considerable and white to elevate do for itself: did seek to opinion expert his tently low. He offered (only leaders political of three testimony statistically sig- was polarization that this knowledge of had first-hand one of whom undermined Dr. fatally it nificant and that dispute) above hotly most the districts analysis. put State probit Epstein’s the effect regarding evidence empirical testimony to refute this expert on no on African declining BVAP claim; its own nor did it submit damning if Even it were opportunities. electoral data on which the analysis of the statistical change the theo- judge for a permissible Instead, using relied. Engstrom Dr. (with notice to no plaintiffs case ry of analyses put to- non-expert decidedly side) evidentiary then make the other asked lawyers, Georgia by its gether subscribes, the party no to which findings data, Engstrom’s but accept Dr. court conclusion wrong is still dissent expert had the conclusions ignore testimony Why? Because it reaches. offer, This which data. drawn way can in no focuses which the dissent credulously accepted, the dissent now that its proving burden carry the State’s adrift from cast itself the court to requires gen- than not likely more plan will Senate analyses and instead expert’s data and the retrogressive effect. no erate and untested inexact its own rely to ^ ^ of cross- significance impressions on the case, voting patterns. important over Such amateur theo- most accomplishes to this rizing entirely inappropriate. The sim- what the State’s statistical data could not. ple controlling fact in this case is that appears The dissent’s if view to be that Engstrom’s expert Dr. analyses data and well-respected few African American lead- expert have not testimony been refuted ers challenged plan, are satisfied with the from the State. pass § it Respect- should muster under fully, this is not the law.

Moreover, while it is true that polar- appear problematic ization trends less I do suggest support not mean to that races, statewide that fact tells us little among minority representatives can never about the elections that are at issue in the be relevant to whether a proposed reap- plan. plaintiffs It Senate was the burden portionment plan comports with the Voting to demonstrate that the statewide data case, however, Rights Act. In this such Epstein. rehabilitate Dr. It did not do so. evidence does Georgia’s not save otherwise The dissent suggests benign now several plan. infirm only For not is the testimony explanations why may white crossover Lewis, Congressman and State Senators greater in regional races. This is Brown and largely Walker irrelevant to' all good, post well and but such hoc efforts court, legal but, issue before the inso- aby judge disguise cannot the fact that relevant, far as it is it does not contradict presented the State not a shred of evi- the statistical data that the dissent seeks dence—statistical or sup- otherwise—-to supplant. to port suggestions. such With bearing proof, the ultimate burden of I am First, any nowhere do of these esteemed simply unwilling gamble to that the serious politicians purport compare to pro- racial bloc undeniably oc- posed with the existing appor- curred in numerous local elections tionment Accordingly, scheme. while disputed districts will not continue under some of they say what have to upоn bear the new plan. (albeit terms) only in the general most Ultimately, we must decide this ease on opportunities available to candi- the basis of par- the record made plan, dates under the new their testimony ties; it is not our role to engage in idle simply does not address retrogression, speculation be, about what might or what which, as we explained, been, could have in lieu of examining and relevant legal inquiry § under 5. Nor do evaluating actually what has been prov- legislators these polarization address the and, more importantly, what has not en— problem that is at the heart of the Court’s proven. been The State has not met its deny preclearance. decision to Their proof burden of on the crucial issues of statements therefore cannot refute the de- patterns polarization. crossover Judi- presented tailed evidence by the United speculation cial cannot overcome this reali- States regarding the effects of this crucial ty. phenomena in places it where matters

most. If the positive lack of polari- racial In an zation data apparent gap effort was the rescue the the center of case, State from the weakness of its the State’s presented by own statis- evidence evidence, tical these suggests dissent estimable men does not come close testimony filling the three African Ameri- void. Whatever the dissent legislators, can only one of whom is actual- has taken testimony from their can there- ly qualified speak about the districts fore be but marginally related to the task party. But that Democratic sans of the Voting Rights by the court to this assigned politicians African American Georgia’s Act. Demo- state safer for to make their sought nei- Indeed, legislators, these three (or does not establish cratic candidates nor Senator Congressman Lewis ther they did not doing that in so imply) even the knowledge of direct any had Walker vot- it for African American make worse voting patterns demographics Indeed, enthusiasm Senator Brown’s ers. their tes- Accordingly, districts. contested general reflect a plan the seems to for levels at BVAP general the timony about that Georgia Democrats among agreement preferred candi- American African which preserve reapportionment will present the compete, to opportunity fair have a dates effectively partisan more their interests ques- to the being tangential in addition such consider- any than alternative. While as bearing little retrogression, tion the impermissible under are not ations in Dis- required might be what levels Act, certainly not they are Voting Rights puzzling 2,12, dissent’s and 26. tricts § 5. demands of satisfy sufficient there- vague evidence to this genuflection rely on the also tries to The dissent result it claims. support the does not fore Thomas that testimony Regina of Senator of Senator Moreover, testimony hardly It is expects to win reelection. she 26, leaves Brown, represents who seeking reelec- politician that a surprising primarily speaking was doubt that he little view that she optimistic expresses tion Democrat, in advanc- loyal interested as a good no It would do her her race. will win party. his own political fortunes ing is lose. What that will to announce she drawing actively involved he was As however, that Senator noteworthy, is more district, his for his home lines proposed the new Senate against voted Thomas shape is unremarka- that new support retrogressive. it is plan, arguing that however, is telling, more isWhat ble. Furthermore, Thomas’ assess- Senator that his ready recognition Brown’s Senator of reelection regarding her chances ments partisan. “[0]f primarily were motives Voting Rights Act irrelevant. currently serving course, I am since incumbents; it protect not votes does I think the I at where looked district thus a dan- minority voters. It is know, protects would, you enhance would politician’s business to conflate gerous the likelihood terms of districts electoral continued of her own it assessment to retain as would be able Democrats protection with the prospects genuine Exhibit seat.” Defendant’s plan Senate strength. American “I African acknowledged at 32. Brown predominant as a looking at race was opinion dissenting Finally, insofar “I am concern,” instead that id. at but evidence legislative this turns to objective that I have looking at the overall it less confused assumption that is per- Democrat to maximize always wanted evidence, statistical contradictory than the formance,” id. at inven- of its own prophet a false it chases whom the politicians on The three only tion. Brown was the that Senator

Given slice of represent but a small relies from a dissent Senator American state African the at- regarding testimony presented on the who testified district contested po- Georgia’s African titudes of behalf, especially sig- his candor State’s to the leadership litical a number of his he and That nificant. Indeed, simply inaccurate to it is plan.

peers support spoken leaders those parti- suggest politics by bespeak sound may well single with a voice. The can Georgia. United States American leaders in Our testimony offered the of a dissenting colleague’s number of effort to convert Americans, prominent African opinions legally from each these into a sufficient basis of the three Senate Districts that it for approving disputed State Senate challenged, plan in which entirely unconvincing. those witnesses ex- is therefore Indeed, pressed quite strange considerable concern about the ef- it is to find the fect changes minority opinion of the on dissenting hinged political to a voting strength. plan, judgment very political for whose nature received, support whatever it has cannot makes it an inaccurate and unreliable indi- fairly represent be said to very thing unanimous cator of the that the dissent preferences or predict. desires of the African would it to use The dissent not evidence, American leaders in the of Georgia. only State relies on this dubious but reality simply complicated more seeks to absolve this responsi- court of its than our dissenting colleague suggests. bility judging by blindly of to deferring judgment very politicians of the whose ac- In the face heterogeneity of such charged tions we have been with scrutiniz- views, it entirely inappropriate seems ing. say, does, as the dissent opinions that the Judge opinion is, of some African pale impor- Americans Sullivan’s for the court view, compared my tance to those correct in its of others. In statement and law, application determining likely impact pro- in its finding correct facts, posed I changes, weighing can see no and correct in its principled upon conclude, instance, basis which to vаrious forms of evidence that have been presented that Congressman Lewis’ to this court. conviction that plan will inure to the benefit of African OBERDORFER, Judge, should, Americans in the context of this concurring part dissenting part. § 5 litigation, given probative more weight than opinion Senator Thomas’ that I pleased join am in Parts III.C.2 and it will not. Nor does our dissenting col- Judge III.C.S of opinion. Sullivan’s I league supply any Instead, such basis. he agree that has met its burden of simply declares that those witnesses whose proving, by a preponderance of the evi- testimony accords with dence, his own conclusion the proposed Congressional probative, credible while simulta- and state House redistricting plans have (or neously discounting ignoring altogeth- neither retrogressive purpose nor effect. er) testimony of others whose views However, respect to the state Senate diverge from his own. plan, I redistricting give greater credence political to the expertise and motivation of sum, then, political opinions Georgia’s political African-American lead- which simply the dissent relies cannot bear ers and reasonable inferences drawn from the weight placed upon has been testimony their data and Why? them. Because these individual ex- statistics than to what I find to be flawed pressions support purport do not even opinions experts conflicting lay wit- seriously address the issues of crossover testimony presented ness by the Depart- voting patterns, and, polarization, most ment of Justice and intervenors. significantly, retrogression, which are the subjects And, § litigation. matter, of this As a legal persuaded I am not addition, political the cited opinions do not that a that reduces the “probability,” conclusively reflect 81-88, the sentiment of Afri- see ante at candidates expert and Department’s testimony of the Georgia’s in three of prevail choice will of districts, lay preserves an witnesses. yet Senate fifty-six candi- for those opportunity equal or fair primary two sources Georgia offered has state- minority candidates other dates that argument its African- support for “retrogres- wide, reason alone is that for opportunity enjoy equal an Americans will §of 5. in violation in effect sive” barely over where win even districts to population is black: voting age of the 50% the is whether us question The before Lewis, John testimony Congressman from whole, has the plan as proposed Senate Brown, and state Robert state Senator abridging denying or effect of “purpose or Walker, majority leader Charles Senate or of race on account to vote right the expert Epstein’s witness and Dr. David To resolve § 1973c. U.S.C. color.” 42 notes, testi- Judge Sullivan the report. As ex- the must determine we question that this case mony expert witnesses minority which, Georgia’s if any, tent to reports fo- expert no unhelpful, with been “vot- retain effective likely to voters are of whether question cusing “on they pos- to what strength equivalent ing” retrogres- are redistrictihg plans proposed plan. Senate the benchmark sess under testimony of Lew- Ante at 64. sive.” Sullivan, it, puts Judge my colleague, As is, Brown, African- and Walker —that must “fact- question analysis that elections can win American candidates closely intensive,” us to examine requiring Georgia where the state of anywhere in changes in which “the context slightly majority, or even are a bare blacks we For task Ante at 76. occur.” will my to less, population voting age of the —is law; jurors only judges of the like entirely inadequately credible mind our responsibili- of fact. It is arewe triers Department’s expert wit- by the refuted effect, and weight, the ty to determine con- from the lay witnesses three ness evidence, the credibili- the value of districts. As trier state Senate tested judge areWe of the witnesses. ty it more fact, find that I therefore would evi- “just any as other testimony expert that: likely than not it, or dence”; reject it or may “accept we it think weight [we] much as it give (1) in- proposed Senate will may draw jurors, we Like deserves.”48 the number to thirteen crease from twelve or con- “any inferences the evidence from Afri- in which districts Georgia Senate sense and common reason clusions that of the vot- majority are a can-Americans a fact To make.”49 establish lead [us] population. ing age the evidence preponderance (2) African-Americans plan, Under must in essence of that fact proponent power elect .retain the statewide will than it is more so likely us “that persuade eleven state Senators. fact determining whether a not so.”50 (3) local statewide and Results preponderance established has been districts, disputed in the three elections evidence, all the consider we “should Brown, Lewis, context of fact, viewed regardless upon that bearing evidence in- testimony and reasonable may and Walker it.”51 produced of who therefrom, establish over ferences its support find for case therefore Id. 2-8. Jury 50. Civil Instructions 48. Standardized. for (1998). Columbia at 3-3 District of Id. at 2-9. Id. at 2-3. next decade there sufficient redistricting will be and on the Senate committee. Brown, increasing support white for can- represents See id. Senator who disputed didates of choice districts Senate District as the served vice- and statewide. chairman of that committee and chaired responsible drafting the subcommittee 4. Black voters in plan. See id. at 41. Districts in tandem with reliable crossover, white will necessary have the participation In addition by individual voting strength to continue to elect their legislators, voting support the of African- of candidates choice. Senate District legislators for all three of the may African-American voters not be proposed plans overwhelming, was able to defeat the white to elect incumbent the case plan, necessary of the Senate their candidate of choice to the Senate adoption. its No of the House member seat, but will retain an influential role in Legislative Senate Black Caucuses voted elections, representing change no from the against proposed Congressional reap- the quo status in terms of that district or in portionment plan. See id. at 43. One terms of the number of candi- House, African-American member of the dates of choice elected to the Senate state- Pelote, area, Dorothy of the Savannah wide. one African-American member of the Sen- Accordingly, I am persuaded that the ate, Thomas, Regina also of the Savannah proposed plan Senate has neither a retro- area, against proposed voted the House effect, gressive purpose nor and is entitled plan, passed in which both chambers with preclearance § under 5. a mаrgin. comfortable Rep- See id. at 46. resentative Pelote and Senator Thomas Findings I. of Fact were also the legis- African-American Although the facts are well stated against lators to vote proposed Senate Judge comprehensive opinion, Sullivan’s plan. See id. at 55. Although the House findings I make and the I inferences approved plan by the Senate a 101 to 71 draw on the basis of the extensive record vote, plan approved was in the Senate require repetition this case some vote, id., by a narrow 29 to 26 see essen- reevaluation. tially political party on lines. See E. John- Dep. son at 27. Ten out of eleven African- Reapportionment A. Process American senators in support voted Georgia’s legislators African-American redistricting plan. support Without the key figures were in crafting the Congres- at least nine of those African-American sional, House, and, particularly, the Senate senators, would have failed. reapportionment plans. Congression- al reapportionment plan was drafted Redistricting B. Senate committee, conference consisting of three state Senators and three members of the Comparison 1. the Benchmark state House. Three of the six were Afri- Proposed Plans can-American. See ante at 42. Of the twenty-nine members of the state House Under the redistricting plan, benchmark responsible committee for redistricting, six Georgia’s fifty-six thirteen of Senate dis- were African-American. id. at African-American, See tricts are majority mea- Twenty-four Senate, members of the state population sured both total and voter African-American, six of whom are registration. served are African-Americans percentage under in their BVAP cline age population voting majority of a con- plan,54along with Senate proposed ante at 55- See those districts. twelve regis- percentage drop comitant African-American. who voters tered 2000 cen- plan, based proposed de- Despite this overall Def.Ex. 118. See dis- data, creates thirteen also sus arguably crease in overconcentrated ma- are a African-Americans where tricts majorities, Department Jus- If voter population. of the total jority three narrowly on the has focused tice yardstick, is used registration that would have districts than half more of districts where number pro- under lowest BVAP levels are African-American registered methodology, Using Georgia’s plan. posed However, voting age if eight. drops to plan Senate District under measure, is the population relevant 50.31%, of Senate have a BVAP would thirteen districts creates proposed plan 50.66%, 12, and a BVAP of Senate District a fifty percent, than greater a BVAP with 26, of 50.80%. a BVAP District one district.52 increase of net Statewide, plan the Senate Testimony African-American of previously “unpack” districts would Leaders Political Afri- voting-age high concentrations had can-Americans, fifty-five from three of its ranging offered from Georgia has and percent, testimo- eighty political more than leaders percent to African-American majorities to black those black under the ny would reduce fraction) (and fifty candidates, the candidates of presumably neighborhood voters, plan, will majority the benchmark black percent. Under choice for sixty in districts ability twelve relevant Senate to be elected two of the retain the only voting-age a 55.43% Afri- majority with a narrow districts —District with can-Americans, BVAP, a 54.73% Districts District with such as Senate and percent sixty less than and 26. BVAP —have dis- benchmark Senate Five other

BVAP.53 Congressman Lewis John sixty and have a BVAP between tricts represents Congressman Lewis Four benchmark seventy percent. District, in the Atlanta BVAP, Congressional Fifth than 70% greater have a districts leadership and area. His metropolitan District 43—88.91% in one—Senate and Rights Movement during the Civil courage See population is black. voting age in the instrumental have been majority- and since twelve 117. Of Def.Ex. these Act and Voting Rights of the 1965 districts, passage a de- experience minority eleven represented by a 12 is methodology, I 53. Senate District which plaintiff's use the 52.I incumbent, Meyer Bre- von Michael white convincing. at n. 25. See more found infra men, in Senate District while the incumbent methodolo- Department Justice's When Fort, is Vincent African-American. of districts gy employed, the number is majority-minority as proposed plan that are See ante 39 increases by BVAP to eleven. Senate District measured falls 54. The BVAP in plan. Addi- Georgia, BVAP in According the benchmark at 56. 1.81% proposed Senate Dis- 34 is tionally, 50.31% the BVAP in proposed Senate Districts 50.54%, net addition to Depart- represents a respectively, which while trict (by districts majority-minority at number of the BVAP Justice calculates ment of calculations), increases 16.58%. Georgia's in District and 49.53% 49.81% Def.Ex. 118. See vitality. Dep. depend specifics its Lewis continuing See more on of that decades, intimately For he 6-11. has been particular candidate and his or her cam- in, about, involved and informed paign. things The kinds of that are South, unique politics particu- important any campaign, like hard politics, larly Georgia beginning with his work, putting together good organiza- directorship of the Voter Education Pro- tion, on, and so will make a difference. ject credibility 1970. Id. at 12. His But a certainly credible black candidate an advocate for African-American voting good has a of winning legisla- chance rights, Department as the Justice acknowl- State, I anywhere tive seat think edges, “beyond reproach.” is Tr. 2/26/02 with a 50% BVAP. Department at 99. The and intervenors Id. at 18. object that Congressman testimony Lewis’ whole, Congressman On Lewis finds irrelevant, because he does not serve in greater the creation of a number of dis- the state not in any does reside minority majority tricts with a slim of the three contested to be districts.55 Howev- er, preferable Congressman opinion Lewis his to fewer “safe” dis- based experience campaigning “[G]iving on his tricts. power real vot- Congress area, the metro Atlanta ers comes from the redistricting kind of familiarity Georgia politics but on his efforts the State of has made with regions at all levels and in all of the state. plans, these for Congress both and for the spent “I’ve a deal great traveling of time Assembly, General House and Senate.... length and breadth the state.... I don’t need to [W]e create these black en- keep up with what happening is all over politically. claves It ability dilutes the the state. It doesn’t matter whether it’s minority to people, elect more and south Georgia, extreme coastal or north to affect majority.” who has the Id. at 23. I Georgia, try responsive to be in- Senator Robert Brown volved.” Id. at 17. Senator Brown is the incumbent state Congressman that, Lewis testified in his Senator in District 26 and chairs the Sen- pоlitical judgment, black candidates have a responsible ate subcommittee pro- for the

better than winning even chance of a dis- posed redistricting plans. Senate Like trict with 50% BVAP throughout Georgia: Lewis, Congressman he believes black can- candidate, I think good a a solid black didates, both incumbents and challengers, candidate, would have more than a 50 can be re-elected and elected in a district percent winning chance of awith with a 50% BVAP: percent BVAP in Georgia. [district] Whether or not a black I candidate wins think the incumbents these dis- in a district with that level of BVAP will tricts at these very BVAP levels are in 55. The argue Dep. intervenors Georgia further Con- is at operates 20. The Senate gressman Lewis is biased because it is in his political plane on a different than the United "best interest[] have Democrats remain Congress, Congressman States Lewis re- control.” Tr. at 158. The interve- 2/26/02 personal alizes no benefit from a Democratic misapprehend testimony. nors his aAs mem- majority in the former. It is reasonable to Congress, ber of the United States it is in his prefer infer that he would control of the Geor- personal self-interest to have a Democratic gia party, the Democratic as it is majority Representatives, in the House of be- constituted, now because he considers it more (and place cause it would minority him other responsive to the interests of his constituents Congress) Members of in line to chair a opposition. than is its House committee or subcommittee. See Lew- leadership positions and Senator to mittee speaking specifically shape. But solid majority I think that to continue to serve as open-seat, an Walker question of matter, candidate would practical Senator an African-American leader. But as winning. He or chance of good only reap have a and Senator Walker Brown good organiza- have to have she would majori- Democratic benefits of a continued hard, rea- there’s no but tion and work if they are able to be ty win can’t why an African-American son newly constituted dis- elected from their you And I can tell BVAP.... at 50% tricts, sup- making unlikely they it would nearly unanimous consensus this. The redistricting plan materially port a in the Senate Black Caucus from the chances. Their reduced their re-election would never that voted for willingness to reduce the BVAP levels a belief had that not been been there majorities to bare their own districts by those senators. shared persuasive they evidence that have such Dep. at 30. Brown in their estimates of confidence the incumbent from Senate they willing As political strength with, familiar Brown is Senator political positions own on the stake their opinion testimo- give singularly qualified estimates. accuracy of those about, and white ny demographics jurisdiction. in his own

crossover trends Testimony Expert Witness expressed opinion that Senate He also in addition specifically, 2 and 12 Districts witness, Epstein, Georgia’s expert Dr. district, competi- would remain to his own in all elections analyzed results candidate of for an African-American tive House, to the the state Sen- state proposed plan. See id. choice under the ate, Congress be- States United 40-42. 2001, using probit analy- 1996 and tween Majority Walk- Leader Charles

Senate purports to calculate analysis That sis. er chance of an African-Ameri- the statistical winning election at can candidate of choice testimony from similar

Georgia offered leader, See ante at 64-65. varying levels BVAP. majority Charles the state Senate elections, in those represents a dis- on the results Walker Based Walker. Senator He was of the voters Augusta. Epstein trict centered in estimated that necessary levels for opinion electing that the BVAP chance of or better have an even of choice an African-American candidate open choice in an seat their candidate of electoral suc- opportunity a fair 44.3%, to have and a 75% if is election the BVAP Congressman than even lower cess were their candidate electing chance of or better Brown estimated: Lewis and Senator rises to 50%. of choice when the BVAP state, I feel would “Generally around analysis His assumes id. at See All at a 45% BVAP level.... comfortable in highly cohesive voters are that black plan in districts of the candidates of respective their supporting Dep. at 12. that level.” Walker well above choice, relatively white voters are while in is reasonable so. “This estimate Walker, less Brown and Senator Senator in cross- difference light significant Democrats, support may well who are both voters, typically it is voting; for black over part in be- redistricting in 5%, white voters while for less than a Democratic likely preserve cause it is around generally it elections Senate, general enabling Sen- majority in the state 25, at 17. com- 20%.” Pl.Ex. his current ator Brown to maintain 2; witness, in Department’s expert Dr. cumbent Senate District the 2000 general and the intervenors’s election between Senator Thomas Engstrom, Richard Katz, witness, Republican challenger. Dr. criti- and a white expert Jonathan Epstein’s methodology, cize see ante at 67- recent Engstrom’s data about the elec- only Engstrom but offers countervail- involving Thomas shows that tions Senator evidence, ing statistical which examines willingness of voters to vote for white racially polarized vot- the extent to which African-American, being po- an far from 2, 12, in Districts and 26. ing exists larized, dramatically. varies the 1999 election, special of whites 20.4% crossed voting pref- To in the assess differences candidate, voted a black al- over and voters, Eng- white and black erences of necessarily go their though votes did not categories at three of biracial strom looked Thomas, to Senator who received in elections benchmark Senate Districts of the white votes cast in that elec- (1) 5.0% 12, and 26: elections to the Senate seat Only tion. of white voters crossed (2) 8.9% districts; in two of the three election runoff, support special in over to her returns each district contests for narrowly against which she won a well- (3) office; statewide local elections However, qualified white candidate. in the largest city populous coun- most and/or following year, running an incumbent inty the district. He concluded that there election, general Democrat Senator (or high polarization racial were levels of Thomas won with 43.6% of the white vote voting) low levels of white crossover nearly of the total vote. 80% Judge contested Senate districts. As Sul- noted, however, “Engstrom livan does not None of the exam- Engstrom elections attempt predict polar- the effect of this typical ined involved circumstances. The ability minority ized on the special election and 2000 runoff were to elect candidates of their choice under non-partisan, meaning that no candidate plans,” redistricting id. at party benefitted from affiliation. Addi- leaving it for to determine us the reason- tionally, there were four black candidates able inferences and ultimate conclusions to special and two white candidates patterns light be drawn from his data election, dividing the black vote. In the all the factual circumstances. election, candidate, general the white being Republican in addition to a in a a. Senate District district, heavily piz- Democratic was also a delivery boy za criminal a record. i. Elections for State Senate Dep. See Thomas at 32-33. Engstrom’s analysis looks at three elec- ii. Statewide Elections tions in non-partisan Senate District two partisan: and special non-parti- one a 1999 Engstrom’s analysis of votes cast in in multiple san election which there were benchmark Senate District in statewide candidates, that, matter, including empirical four African-Ameri- shows as an races cans; special non-partisan a 2000 significant runoff numbers of white voters top between the winner and African- crossed over to vote for a black candidate. Thomas, vote-getter, Regina Engstrom analyzed seven statewide races: who won that election to the in- primary general become the Democratic and elec- 56. Statistics in this section are taken from the umn in Def.Ex. Table 1. King's Ecological methodology Inference col- Commissioner; District ev- respect proposed Senate for Insurance tion runoff, past in a ery candidate who ran Democratic primary, Democratic Labor would have carried the for Commis- statewide election election general sioner; primary district, for Public receiving Democratic between 57.4% proposed Commissioner; cast, general and the the total votes a fact and 77.7% of Service In the four Attorney in for General. in a race election from which results Senate inferred, races, crossover for Democratic-only white reasonably may new district from a low ranged candidate the black iii. Local Elections 31.8%, in runoff for Labor Commis- Savannah, largest city Local races 58%, primary for sioner, in the high to a District show low levels of Senate In the Commissioner.57 Public Service voting, ranging from 2.8% white crossover elections, black Democrats where general analyzed eight elec- Engstrom to 10.6%. Republicans, 27.8% against off white faced tions, involved two but those elections district voters in benchmark of white Additionally, Engstrom black candidates. Insurance candidate for voted for the black codes all of the local elections Savannah 44.9%, Commissioner, re- 44.8% and non-partisan, atypical. Eng- which is the black candidates voted for spectively, analyze any elections sur- strom did not Attorney for Labor Commissioner rounding County. Chatham General.58 of these results probative value b. necessary reliance by Engstrom’s affected i. Elections for State Senate dis- conducted benchmark on elections analyzed the 1996 and 1998 Engstrom differences significant There are tricts. state Senate primaries Democratic for the political geography between 12. The black race benchmark District districts, proposed due to benchmark choice, White, long- John candidate precincts and the of some the removal member, narrowly lost both time House ante at 85-86. To of others. See addition white primaries, each time to different this, performed a re- Engstrom control 1996, only 10.6% of white candidate. examines vot- analysis, which aggregation Al- support voters crossed over to White. precincts in the that consti- ing patterns bid, he lost his 1998 Senate though also See Def.Ex. tute the district. respect- climbed to a more white crossover analysis re-aggregation 4. His Table able 17.5%. percentage of the vote only the total shows ii. Elections Statewide a candidate would receive analysis of the statewide district, Engstrom’s breaking down without plurality in District shows voters. races of black and white preferences that, to vote for black crossing over However, with white voters analysis indicates candidate for Attor- Democratic-only 58. The African-American four 57.The results of the General, Baker, disprove hypothesis that some white prevailed in ney races Thurbert they politically partisan that voters are so represents the state of Geor- that election and over a support a black Democratic would litigation. gia in this they Republican, also so racist that white but if a a black Democrat would never vote for taken from the in this section are 59. Statistics running. Smith v. Democrat were white Cf. methodology col- Ecological King's Inference Allwright, 64 S.Ct. 88 321 U.S. 611, Table 2. umn in Def.Ex. (1944) (holding white-only prima L.Ed. 987 Adams, unlawful); Terry v. 345 U.S. ries (1953). 1152 S.Ct. 97 L.Ed. in ny, partisan which he described as some primaries In thé Democratic candidates. runoff, ranged years non-partisan others. crossover 33.5%, primary in the Labor Commissioner races, white crossover In the House (where there were more than two candi- ranged from 9.9% to 17.1%.61Three dates), in the Public Service Com- to 58% four House races involved the same black primary. missioner For the two remain- candidate, Roberts, while White ran John races, Democrat-only whites ing 44.1% of county-wide race. In the two the fourth for Insurance voted for the black candidate Dougherty County races in —Democratic Commissioner, and 40.8% voted for primaries county for the chair of the com- black candidate for Labor Commissioner county mission and for coroner—white in the runoff. 21.3%, respective- crossover was 9.4% and elections, the black candi- general

In the ly- for Insurance Commissioner received date analyzes mayoral four races Engstrom However, cast. 23.8% of the votes white Albany: partisan general elec- candidate for Labor Commission- the black tion,62 primary, the 1995 Democratic votes, in race er received 48.1% of white election, general and the non-partisan Republican, against a white and 44.6% of non-partisan general election. In the whites crossed over to vote for the black primary, Engstrom estimates white Attorney candidate for General. crossover votes for two black candi- These results are based on votes cast 4.8%, stronger dates totaled with the can- the boundaries of benchmark Sen- within race, receiving 4.0%. In the 1997 didate Engstrom’s re-aggrega- ate District 12. white crossover for the black candidate analysis, which cast tion looks votes was the four black candi- 5.4%. within the boundaries of the dis- dates the race received 12.8% of white trict, shows that the black candidate for cast, strongest votes with the candidate nar- Insurance Commissiоner would have receiving though support 9.7%. Even white lost, rowly but all other candidates but one for black candidates remained under 10% majority would have received a of votes mayoral Albany, elections cast.60 support throughout the 1990’s white

iii. Local Elections steadily black candidates increased. respect Engstrom With to District c. District 2663 analyzed types three of local elections: i. Elections for State Senate seats, partisan parti- races for state House positions Dougher- san races for elected There have been no black-white contests ty County, mayoral and the race in Alba- for the seat in District 26 under the primary In the multi-candidate race for 62. The 1993 election is outside the normal Commissioner, race, Labor the black candidate scope Engstrom’s analysis. In that votes, would have received of the more 42.6% Republican black candidate ran as a re- any primary than other candidate in the but ceived of white votes and 3.1% 42.1% majority. less than a election, black vote. In that the white Demo- opponent preferred cratic was the candidate 61. John White received of the vote 17.1% by both black and white voters. when he ran for election to the House in 1994, predating negative publicity he re- forming lobbying compa- ceived in 1996 for 63. Statistics in this section are taken from the ny political to trade on his connections. This King’s Ecological methodology Inference col- suggests may white crossover 17% umn in Def.Ex. Table 3. White, typical for candidates like primary 1996 Senate aberrational. Brown candidates each received 34.2% of white plan. Senator current benchmark votes, in elections held in non-partisan general in a 1991 1994 and initially elected was election, A African-Americans 1998. black candidate for district attor- special when voters, ney in registered gar- received 16.6% white votes were a in primary. to defeat a credible Democratic But the Demo- nering 56% of the vote county primary ante at 62. cratic for the Board of opponent. white See Commissioners, only 2.7% of white voters ii. Elections Statewide supported the black candidate. In all seven state-wide races between Macon, Engstrom’s data shows that candidates, in and white white voters support for three different black white District 26 demonstrat- benchmark Senate city council at-large candidates for seats support candi- ed considerable for black ranged general from 14.2% in the 1995 races, Democrat-only In the dates. election, general to 22.9% in the 1995 elec- for Labor black candidate Commissioner tion, in general to 27.4% the 1999 election. votes in the multi- received 33.6% of white However, in mayoral the 1999 Democratic primary, and in the two- candidate 36.9% primary, only supported 10.4% of whites The black candidate for person runoff. the black candidate. garnered 41.5% Insurance Commissioner cast, of white votes and 57.4% of white supported the black candidate Testimony Lay of Witnesses Commissioner. Public Service Department present- The of Justice has elections, general In the 32% whites declarations, testimony, in the form of ed for the black candidate for Insurance

voted public local officials and oth- from nineteen Commissioner, 46.9% voted for the black who in the community er leaders reside General, Attorney and 54.9% candidate for disputed Senate districts. See Def.Exs. candidate for Labor voted for the black face, 501-519. On their these declarations re-aggre- In Engstrom’s Commissioner. paint grim picture racially polarized analysis, looking voting pat- at how gation districts, in voting the three Senate play terns would out Sen- ability cast doubt on the of African-Ameri- ate District black candidates received in cans to elect of choice each of candidates in majority every two-person of votes However, the the districts as redrawn. race, than more 60% of received probative value of this testi- credibility and In the vote five out of six races. mony seriously compromised signifi- primary, candi- multi-candidate the black cant between these stark contradictions top vote-getter, with 49.6% of date was the declarations, testimony the more nuanced support. white when cross-examined in of these witnesses iii. Local Elections the actual results in depositions, their local elections. Department’s expert analysis of lo- District 26 ad- cal elections a. Senate District partisan County elections in Bibb

dressed and Macon. White crossover testimony Department offers of ten gen- local elections Senate District 26 is district, this all of whom witnesses from than in either erally higher polarized voting pre- will racially aver that succeeding vent black candidates 2. The most im- County proposed Senate District In races for seats on Bibb Education, portant testimony comes from Senator two different black Board optimistic any the chances of in less about the incumbent Thomas.64 As successor, pat- “[pjeople may not know on racial because her views any name rec- terns, potential They may and the not have potential and her them. part most a of choice to be elected and I think for the ognition, of future candidates district, because, thought- I proposed in deserve would not win said oppos- in tune Thomas are more ful consideration. Senator earlier whites it plan because removes whites.” Id. at 122. es the Senate district, despite its of her current parts Department presented also declara- who underpopulation, particularly voters African-American tions from four Savan- her; also previously supported had Goodman, Gwendolyn nah aldermen: County among it divides Chatham because Jackson, Jones, Jr., Branch Clifton Edna present of the two. three senators instead Two African-American and David Jones. Dep. at further testi- Thomas 82-83. She Commissioners, County Harris Chatham understanding fied to her Rivers, Odell, Murray also Jr. Joe 50%, a under Senate District has BVAP opposition plan. in to the Addi- testified you your “when have generally, and that tionally, three local citizens who are active percent voting age numbers below against testified members NAACP population then nine times out of ten plan: Dr. Prince Jackson is former representative a white you’re going get president of the of education vice board Id. at 102. Senator Thomas senator.” president of and a former Savannah State perceives that her numbers are low- BVAP acting is the College, Richard Shinhoster senators, than other African-American er president of the Savannah branch of the minority vot- unfairly disadvantages which NAACP, and Helen S. Johnson is the CEO district, in testified she would ers her but rights of the local civil museum and if opposed even her numbеrs member of the Executive Committee of equivalent, to the addition of a were due the Savannah branch of the NAACP. county. third senator in the See id. at 125. community lead- These African-American her nar- Senator Thomas concedes that fairly confident ers Senate District margin victory special row the 2000 progressive the more voters in white election, votes, which she won 70-odd neighborhoods Savannah’s historic will likely racially was for reasons other than candidate, continue to vote for better polarized voting. opposed by a She was race, irrespective primarily and are con- candidate, popular, well-qualified white heavily cerned that the addition of white Braun, previously Dana who had served as heavily Republican precincts from the at-large city an alderman in the of Savan- islands area could cause Senator Thomas substantially nah. Braun was better-fund- by Republican challenger. defeated ed and was endorsed African-American voters, Goodman testified that white ministers, elected officials and while not a ward, primarily on a least her vote single black elected official endorsed Sena- ability. Dep. candidate’s Goodman at 29. 22,142-143. tor Thomas. See id. at Jones, declaration, Clifton his refers to a BVAP, “general prefer rule” that white voters

Despite the reduction Senator ¶ candidates, personally Thomas thinks she can win the white C. Jones Decl. but support district. See id. at 121. She was believes that he has received the provide 64. Senator Thomas did not a declara- See Def.Ex. 704. ties. tion, deposed person by par- but was in in overwhelmingly Tybee the white voters his even white Is- roughly 40% of 11-12. David land “because I have a lot of I Dep. people Jones at know ward. C. Tybee.” Dep. that there are white citizens out Rivers at 42-43. Jones affirms support in who will vote for a black Prince Jackson received white Savannah a candidate. D. when he defeated a white candidate in a candidate over white stated in her 1970 election to the Dep. at 19. Johnson Savannah-Chatham Jones voting County “racially polarized declaration that Board of Education. P. Jackson usually” Dep. ... occur in Savannah at 7-8. He estimates that patterns 20% ¶ elections, Decl. but ex- white voters will typically H. Johnson cross over for a black, “people in plained deposition her don’t candidate and that 70% to 80% of support racial ... in really might very strong vote because of issues white voters a city [They vote] [o]n elections.... candidate. See id. at 62. “I parties, say percent I think the different would people issues and 20 to 30 ... Dep. point H. at 41. haven’t parties.” gotten they Johnson to the where can vote for the other race.” Id. at 58. in There is evidence that black voters political in have been able to form Blanket statements the declarations Savannah city’s questioning electability coalitions with the Jewish communi- Senator Thomas’ in 26; ty. Dep. Dep. proposed significantly at E. Jackson district are qual- See Odell Additionally, African-American ified deposition testimony. at 17. witnesses’ declaration, In candidates who are Roman Catholic draw his Clifton Jones stated support “probably from white Catholics the Chat- that Senator Thomas will not be County Dep. against strong oppo- ham area. See Goodman at able to win a white ¶ 33; Dep. city’s deposi- P. at 8. The white nent.” C. Decl. 21. In his Jackson Jones tion, political although stating establishment has shown some Senator Thomas willingness to advance African-American “would have a better chance under the old candidates, plan,” notably agreed most the endorse- he that she had a fair chance mayoral winning proposed ment of a black candidates district. C. declaration, In long-serving mayor. Dep. white See Jones at 48. his Shin- Savannah’s 96; Dep. “unlikely that Dep. E. Jackson at H. Johnson at hoster stated that it was elections, city Regina In two recent white Senator Thomas will be reelected 2,” candidates made racist remarks about Senate District Shin- ¶ 16, deposition opponents, their African-American but hoster Decl. but his expressed Thom- appeals opinion those were unsuccessful and re- that Senator strong harm the white candidate as “would be a candidate for re- bounded to Dep. Dep. at 28. among white voters. See E. Jackson election.” Shinhoster 80; 24; Dep. D. Jones at P. Jackson Much of the concern about Senator Dep. at 30. chances are based on Thomas’ election great- party politics, rather than race. David terms of white crossover area, pro- County thought clarified that he believes the er Chatham Odell Jones fire,” D. pick up posed district throws “into the Senator Thomas “would her ¶ 5, overwhelming majority “Regina of the lower middle Jones Decl. because runs as [t]hey her in a white put income white in an election in the Democrat [and] voters” Republican they extended her Dep. district. Odell at 25. Riv- district Odell, Dep. D. at 14. who ers testified that he draws white votes district.” Jones County, spec- stated in his declaration Senator Savannah and Chatham majority badly” by beaten garner ulated that he could Thomas “would be ¶ David Williams candidate, representing Ward 3. Decl. Odell strong white member, he believed 6. deposition in for Ward city clarified his also a council win the Democratic Thomas would Senator president of the local Wright a former general in the election primary but lose branch, unsuccessfully ran for and NAACP Dep. at 24. Republican.65 Odell a white county board of commis- to the election county board of education. sioners and of the local depositions, In their several sanguine about leaders were more presents 12 a somewhat District Senate candidates African-American prospects of currently district is unusual situation. The their Thomas than dec- other than Senator incumbent, Michael by a white represented believes larations had indicated. Goodman Bremen, sup- who was not Meyer von politician, such as a well-known black in his by majority of black voters ported African-American herself or Savannah’s of local Afri- election. The concern 1998 Adams, a fair “get would mayor, Floyd politicians in this district is can-American district. in the Senate shot” incumbent, but the retention of an Dep. at 32. Rivers seconded Goodman challenger ability of a rather the opinion that she could elect- be Goodman’s Bremen.66 Meyer defeat von seat, although he new ed to the exceptional candidate. believed she was an defeat attributes his Sherrod Dep. at 47. Rivers voting. defeat to racial bloc See White’s ¶¶ declaration, 6-10. his Sherrod Decl. 12 b. District white voters in stated that “[m]ost Sherrod Department testimony offers simply will not vote for Gеorgia Southwest from Senate District five black declarants ¶ Id. 11. When Senate candidates.” White, Sherrod, Arthur 12: Charles John Afri- explain the success of two pressed to Williams, Williams, and David William major- judicial candidates can-American Albany city Wright. Sherrod is a former Dougherty County, as well as ity-white commissioner, unsuccessfully ran for who ability to re-elected to Bishop’s Sanford the District Senate seat majority-white Congress in a district campaign adviser his acted as White’s Sherrod which includes Senate the same and 1998 election efforts for here in phenomenon fell back on “a down House for seat. served the state White explain that we can’t some- south in the 1996 and years, and was defeated times,” sup- that occurs when white primaries Democratic for the Senate port Dep. at black candidates. Sherrod opponents. white Arthur Williams seat council, Albany city 97. is a member of Although Department's wit- Proposed two of the Senate District would be heavi- precincts ly complained Data all of the Democratic. Senate District 12 nesses from comprise that will district indi- Meyer adequately repre- von Bremen did not supported those voters Al cates that Gore, 64.41% the African-American sent the interests of Barnes, voted for Governor 67.84% University study community, a Clark based on supported than the African-Ameri- more 70% voting records of Senate members found in the elections can Democratic candidates consistently Meyer Bremen voted von supra. for statewide offices discussed See Pi. Caucus. See United with the Senate Black heavily tilted towards Ex.2D. In a district so Response States’ to Plaintiff's Post-Trial Pro- candidates, may reasonably be Democratic it Findings posed of Fact and Conclusions of likely to face inferred that Senator Thomas is ¶Law 489. challenge primary than in the a stiffer general election. in District the declarants’ concerns about of his losses explanation White’s

John racially polarized voting are contradicted in both his declara- primaries, the Senate by the success of their own candidacies. focused less deposition testimony, tion and political factors. and more on on race that, personal Abrams testified his n White the current Lieutenant Gover- faced experience, patterns polarized “voting nor, primary, Taylor, the 1996 Mark County, racial in Bibb Abrams along lines” him in the substantially outspent Taylor ¶ acknowledged deposi- Deck 6. He in his ¶ Decl. 12. White campaign. See White personal experience tion that his includes race, publicity in that adverse also suffered elected, being sup- with substantial white newspapers local and Atlanta when the in port, against qualified a white candidate company that he had founded a reported Dep. at county a with 43% BVAP. Abrams capital- called “Connections Unlimited” a white Democrat 15-22. Barnes defeated 22-year service in the state ize on his primary Republican in the and a white in result, at- Possibly House. as White election, general winning nine out of voters, more ¶ tracted few crossover 2. In precincts. ten Barnes Deck Hart’s supported of black voters his than a third bid, him, only election whites not voted also cited racial opponent. white White sponsored campaign also events for but primary, see appeals by Taylor homes, although him in their he is not ¶¶ 12-18, harm quantifying id. without enjoyed certain he would have the same campaign. City to his council member if support opponent his had been white. that lost suggests Arthur White Williams Dep. Hart at 24-25. Taylor was better-funded 1996 because Testimony from these witnesses indi- “the race card.” A. played Williams likely cates that it is that Robert Brown ¶Decl. 6. in the redrawn district. can be elected Hart stated that Brown “is a shoo- primary,

In the 1998 which featured no Senator distractions, winning to in for re-election in Sen- such White believes he lost ¶ 9, 26,” Hart Deck and other Meyer von Bremen because of weak turn- ate political agree that figures voters. “I think some local Senator among out black of good out to Brown retains at least a chance people did not turn vote be- ¶ 7; Deck being I elected. See Abrams they figured cause was sure bet ¶ ¶ ¶ 7; Barnes Deck Ellis Deck 7. The local win.” Deck White political leaders are more concerned c. Senate District 26 compromise district will redrawn African-American candidates ability of Department Five witnesses for the of- other than Senator Brown be elected. testimony opposition pro- to the fered minority “a non-incumbent Ellis believes posed plan pertains as it to Senate District chance candidate would 50/50 Abrams, the African-American 26: Albert of-winning proposed Senate District 26.” president County the Bibb Board of ¶ Ellis Deck 9. III, Education; Afri- Barnes an William However, testimony, deposition their County of the Bibb can-American member Education; III, political leaders from Senate District Bert Bivins an local Board of pool identified a of black candidates county commissioner for 26 African-American Ellis, Robert Brown. the African- who could succeed County; Bibb C. Jack Ellis, Macon; as sever- agreed F. Abrams as well mayor American and Samuel members, Hart, city African-American council county an African-American com- al any opt- if be formidable candidates County. for Bibb Senate would missioner spect to their exercise of the elec Abrams ed to run for the state Senate. effective Beer, an African- Dep. at 61. Barnes believes 425 U.S. at toral franchise.” added). candidate could be elected from Neither the (emphasis S.Ct. with a 50% BVAP. a Macon-based district § in 5 nor authoritative decisions text pool Dep. at 59. Hart described a Barnes require preservation it terpreting candidates potential African-American majorities that would super “robust” Abrams, could, like “draw across who candi guarantee election as County, Dep. Hart board” Bibb choice; precedents the statute and date attract level of potentially the same well minority’s op “merely ] mandate[ that Senator Brown en- white crossover representatives of its portunity to elect 1991 election. Id. at 51-52. joyed his diminished, directly or indi choice not be rectly, by the actions.” Bush v. State’s Analysis II. Vera, 952, 983, 116 1941, 135 S.Ct. U.S. jurisdic- that certain requires Section 5 added). (1996)(emphasis L.Ed.2d 248 them, tions, pre- Georgia among obtain necessarily equate Opportunity does not Department from the of Justice clearance three-judge panel probability, although majority a of the United States so or to 81, the District of Columbia District Court for Each ma- holds. See ante at 82-83. implementing any change (and a cases, before jority-minority district some standard, prac- “qualification, prerequisite, minority popu- districts with a substantial tice, procedure” respect voting, or rеpresents an majority), lation less than a reapportion- including redistricting opportunity minority for a candidate of ment, proposed change to ensure that the majority, choice to be elected. The rather purpose not and will not “does have comparing majority- than the number of denying abridging have the effect of or the minority or the of minori- districts number or color.” right to vote on account of race likely of choice to be elected ty candidates § 42 1973c. U.S.C. proposed plans under benchmark determining whether is enti- opportunity, as the measure of look to the 5,§ preclearance tled to under we must probability minority that a choice candi- on the of the facts “wheth- determine basis district, specifi- date will be elected each ability minority partici- groups er the of cally minority the chance that a whether pate political process and to elect candidate of choice will be elected has [ajugmented, their choices to office is di- decreased from a “robust” chance to “a minished, change or not affected ‘reasonable’ or ‘fair’ chance.” Id. at 77. affecting voting....” H.R.Rep. No. 94- legal authority There is no for the ma- 196, States, 60, p. in Beer v. quoted United § jority’s proposition requires 5 that a 130, 141, 1357, 425 U.S. 96 S.Ct. 47 plan preserve pre-existing probability a (1976). L.Ed.2d 629 There is no retro- prevail. that a choice candidate congressional as defined gression, Court, contrary, Supreme To the albeit committee and reiterated in Beer and sub- context, consistently § held opinions, sequent redistricting where a Rights provide that the Act aims to Voting plan augments or has no effect on the opportu- nothing equal more than a fair or voting power minority group; of a retro- nity, guarantee “safe” if does seats gression plan occurs diminishes position victory.67 “the of racial minorities with re- or a “robust” chance of Other purpose Voting Rights provide § Act In the context of it is clear that the of the § 5 or increases the number of districts state recognized, courts lower in- context, preserves or that a wide in which minorities have a fair or where mi- number of districts creases the opportunity reasonable to elect candidates equal have an or reasonable nority voters preclearance, of choice is entitled to candidates of to elect their opportunity every whether district must remain at or Colleton retrogressive. is not See choice improve probability on the benchmark of McConnell, 01- v. No. County Council victory, doing even if so maintains a minor (D.S.C. 3581-10, op. at 95 Mar. slip ity super-majority far in excess of the level court) 2002) (examining the (three-judge needed for effective exercise of electoral majority-minority districts number City franchise. Richmond v. United Cf. opportuni- equal “at a level of maintained (in States, 422 U.S. at 95 S.Ct. 2296 Byrne, v. 740 F.2d ty”); see also Ketchum annexation, § context of 5 does not Cir.1984) (7th (defining retro- num require the maintenance the same decrease in “the number of gression as a minority-controlled city council ber have a reasonable wards which blacks seats, doing so “permanently when would to elect a candidate opportunity voters). overrepresent[ minority How ]” choice”). § does not conflate a This ever, in v. Mississippi, United States Rather, § it inquiry. inquiry with U.S. 100 S.Ct. 62 L.Ed.2d 739 simple comparison recognizes that (1980), Supreme summarily ‍​‌​‌‌‌​‌​​​​‌‌​​‌‌​​‌‌​‌​​‌‌‌‌​‌‌‌‌‌​‌‌‌‌‌​‌‌‌‌​‍af Court un- majority-minority districts number by three-judge panel firmed a decision a of proposed plans, der the benchmark States, court, Mississippi v. United this employed by the although traditionally (D.D.C.1979) F.Supp. (Wilkey, 569 490 courts, by because it is itself insufficient Greene, JJ.),68 pre- granting Pratt and question of whether the fails to answer majorities Mississippi’s reapportionment at a level that enables to clearance “ef- franchise,” of the electoral exercise Al plan, dissenting. Marshall with Justice fective (em- Beer, 141, 425 U.S. at 96 S.Ct. 1357 retrogres lower court found no though the added). phasis plans had the sive effect because “both Negro same number of districts with vot majority Supreme A of the Court has more,” 444 of 60% or ing-age populations of definitively question answered the never 1058, 994,69 100 Mar- redistricting plan preserves a that U.S. at S.Ct. Justice whether (1993) (“Only apportionment "equal” opportunity. "The most 500 if the an or "fair” protected reading language sug- denying a that would scheme has the effect of natural opportunity gest equal 'opportunity’ equal to elect its candi- that citizens have an class the 2.”). § participate process an does it violate to in the electoral date of choice representatives equal 'opportunity' to elect Greene, Judge ap- they given free and Harold as chief of when have been the same Rights pellate of the Civil Division of open access to the ballot as other citizens and section Justice, principal properly Department counted. The was a their votes have been Voting Rights well as speaks opportunity an draftsman of the Act as section in terms of —a elect, Rights participate and to an the Civil Act of 1964. chance'—-to ability any particular re- assured to attain Hall, 874, 925, finding three-judge court made a v. 512 U.S. 114 sult.” Holder (1994) (Thomas, voting-age population “Negro was L.Ed.2d 687 a 60% S.Ct. 129 J., Johnson, Negroes necessary to have a concurring); v. in order see also Miller fair 900, 927, opportunity a candidate of their 132 to elect 515 U.S. 115 S.Ct. (1995) (the 994 provide 444 U.S. at 100 S.Ct. Act seeks to choice.” L.Ed.2d 762 added) (Marshall, dissenting). gain public (emphasis J. "equal opportunity to office re- Quilter, race”); disagree did not with the fair gardless Justice Marshall Voinovich v. analysis, thought BVAP opportunity but L.Ed.2d U.S. 113 S.Ct. (2000) 866, 145 L.Ed.2d 845 have held that the S.Ct. nonetheless would shall //”).70 (“Bossier “In vote dilution retrogressive had a effect Parish proposed plan the BVAP is certain backslid- prevents nothing § it decreased but because cases County § from 71.72% in Leflore 5 affirms pre-clearance districts: under ing, 64.26%, County Marshall from 62% backsliding.” nothing but absence 56%, County from 70% to and Adams Id. 6,100 & n. S.Ct. 994. 63%. See id. principles It from these follows Marshall would have re- Although Justice where, here, retrogression no there is majori- maintenance of BVAP quired the “sidesliding” rather the facts establish than at percentage, rather ties at a certain “backsliding.” precedent No ad- than election, there a probability of is a certain “sidesliding,” kind of occur- dresses this higher the two:- BVAP linkage between majority here the ring where as BVAP incrementally higher an results in at least proves the be less than proposed district ante of electoral success. See probability alignment but new prefiguration, Mississippi, outcome in at 81-82. The the number of BVAP districts increases binding precedent on this although not level African- statewide and retains the court, squarely re- three-judge nonetheless representation in the Senаte. adopted here the ma-

jects position plan comparison to the may retrogres- found jority, plan that a plan likely more than not benchmark districts, in individual sive due to declines (1) many likely to create statewide as whole, plan as a such as though even districts, as mea- majority-minority more here, maintains the Senate (2) BVAP; to make it rea- sured voting strength equiva- statewide at levels anticipate the number of sonable plan. lent to the benchmark minority candidates statewide successful proving, by Georgia bears burden equal will or exceed the number elected evidence, that preponderance its plan. under the benchmark redistricting plans other Senate and § Voting Rights consistent with 5 of the Retrogressive A. Effect Act, A burden is a limited one. but its yet Today, unprecedented, we face an redistricting plan may pre-cleared be un- unforeseen, challenge:. not to assess 5, “enjoined § yet der still be as unconsti- a deliberate reduction of black tutional,” Reno, 630, 654, whether v. 509 Shaw U.S. 2816, (1993), super-majorities, undertaken with the en- 113 S.Ct. 125 L.Ed.2d 511 or legislators § 2 dorsement of African-American result in vote dilution actionable under Act, expectation v. with the reasonable that these Voting Rights of the see Reno Bd., 320, 335, majorities, bare combination with mean- Bossier Parish Sch. 528 U.S. necessary oppor- percentage provide to a fair that ‘result’-—is to increase further the serious tunity in rural counties was in excess of already implicated by § 65%. costs 5.” federalism 1057, See id. at 100 S.Ct. 994. Bd., 471, 520 U.S. Reno v. Bossier Parish Sch. 1491, (1997) 137 L.Ed.2d 730 117 S.Ct. decisions, Bossier Parish the Su- (internal omitted) ("Bossier Parish citation emphasized jurisdiction preme that a Court I”). Supreme Court did not consider the seeking preclearance § under 5 need not plan a converse issue of whether is not prove plan satisfy stringent § that its 2’s more preserves § under because it a vote-dilutive requirements. require “To a vote dilution equal opportunity minority candi- fair or jurisdiction litigate its whether prevail, may dates of choice to nonetheless redistricting plan also has a dilutive ‘result’ implement plan retrogressive § before it can if under —even Attorney proving General bears the burden of

119 districts, voting, perhaps would have er the in- white crossover dilutes ingful mi- preserving enhancing minority the effect of or group fluence of the as a whole. statewide, is none- nority voting strength I, Bossier Parish U.S. it reduces retrogressive theless because (Thomas, J., (inter- concurring) S.Ct. dis- in three contested Senate the BVAP omitted). nal citation fifty percent. just tricts to above an “Retrogression” is often-used but ill- Supreme Although majority term, just not parties defined in this issue, leaving not faced the us Court has case, but the case law as well.71 As individual ground, on uncharted Justices out, Judge pointed Sullivan the num- a situation such as this. have foreseen majority-BVAP pro- ber of districts Supreme When the Court introduced the posed plan “may appropriate be a more Beer, concept retrogression Justice in determining number to consider wheth- “it recognized Marshall will not al- properly er a district is characterized aas ways easy be so to determine whether a district,” ‘majority-minority’ ante at al- plan Negro increases or decreases new though it is not a measure to rely prior voting power plan,” relative to the exclusively. id. judges See Other anticipating situations where the effective- used the number of elected candidates of minority voters could be reduced ness gauge ability choice as a of black voters’ “packing” them into districts where effectively their would “Is it not as exercise electoral fran- their votes be wasted: gerryman- common for minorities to be chise under the benchmark and See, Holder, sepa- into district as into plans. e.g., dered the same 512 U.S. at 895- Beer, (Thomas, J., rate ones?” 425 U.S. at 156 n. concur- S.Ct. (Marshall, J., dissenting). measures, S.Ct. 1857 Jus- These numerical without ring). Thomas, discussing trade-off tice be- more, may may finally determine high percentage tween a of minorities plan retrogressive. or not a whether certain districts and influ- within But, event, any analysis numer- these statewide, stated that: ence necessary predicate ical measures is reapportionment that a have held

We fact the conclusive decision: the effect of position of racial that “enhances proposed plan minority’s on the ultimate by increasing' the number of minorities” ability effectively. to exercise its franchise majority-minority districts does not diluting abridging or “have the ‘effect’ of Defining 1. Effective Electoral right the to vote on account of race Strength § in meaning

within the 5.” But so ways different to define There are three holding studiously we avoided address- majority-minority: by a district as its total necessary consequences ing one of population, by voting-age population, its increasing majority-minority districts: by registration. Under the first mea- necessarily action decreases the Such sure, Georgia’s proposed plan is minority level of influence surround- districts, numerically retrogressive in the sense not ing and to extent “dilutes” Afri- districts where the vote of those oth- number of 71. It retrogres- minority population may to define what or BVAP of an individual be easier fatally retrogressive. example, § deemed See Ket sion is not. For 5 does not re- district chum, suggest, retrogres- (rejecting a similar quire, 740 F.2d at 1414 as intervenors approach approach to district-by-district as "too inflexible an sion be evaluated on basis, any practical redistricting”). decrease in needs of well as statewide

120 (S.D.Ga.1995)); 1556, 1568 n. 18 majority F.Supp. comprise a can-Americans Grandy, thir- 512 U.S. constant at also v. De population total remains see Johnson 2647, 997, 1018, 14, the relevant teen. If BVAP is used as n. 114 S.Ct. 129 measure, plan numerically (1994). is L.Ed.2d 775 - ameliorative, increasing the number of dis- Here, majority I follow the would one, majority tricts with a BVAP voting that have embraced lower courts if voter thirteen.72 But black twelve to ingredient the relevant age population as measure, appropriate registration is the Voting minority voting strength plan retrogressive, is than the courts have litigation. Act Other Rights majority-black diminishing the number of percentage “consistently relied” on BVAP districts five. “In analyz § ante at 79. 5 cases. See expressed not Supreme Court has factor, fairness ing the racial any of three mea- preference for clear (VAP) num population is the relevant age Beer, Supreme found Court sures. determining whether ber be used that increased the number of that a particular in a district will be minorities majority population districts with elect a candidate of their choice.” able to two, and increased the number from one to Wetherell, 1076, F.Supp. DeGrandy v. 794 majority regis- with a of black of districts (N.D.Fla.1992); v. see also NAACP 1084 one, was amelio- tered voters from zero to (5th Fordice, Cir.2001); 252 F.3d 364 Beer, 141-142, 425 at 96 rative. See U.S. (8th Hunt, 1042, 1044 Emery v. 272 F.3d any “expressing] S.Ct. 1357. Without Cir.2001); F.3d Cooney, Old Person v. opinion subject,” the Court reiterat- (9th Cir.2000); 1113, 1122 v. Lib Solomon determining trial court’s dilemma in ed the Comm’rs, 221 F.3d erty County registration voting age popula- whether (11th Cir.2000). three-judge Two other superior Registration tion was a measure: courts, determining whether their court- of electoral greater created a likelihood “ § plans 5 re redistricting ordered met success, ‘in essence condones voter but ” popula quirements, voting age have used Johnson, apathy.’ Abrams v. 521 U.S. v. 74, 94, tion as the relevant statistic. See Smith L.Ed.2d 285 117 S.Ct. (1997) Miller, Clark, F.Supp.2d 540-541 (quoting Johnson v. Harrison, Joseph Department of Justice Dr. Roderick former chief competing proposed two methods for the Racial Statistics Branch at the U.S. Cen- calculating Bureau, With all due deference to BVAP. Department's sus that the allocation Department's guidelines, see Bossier Par- statistics, justified as a matter of rule is I, S.Ct. ish 520 U.S. evidence, considerations, empirical or other Department's voting age definition of black such as communities of interest. See Pl.Ex. population is unreasonable. For the first time, respondents to the 2000 census were methodological dispute of this Resolution identify belonging permitted to themselves as analysis, critical to this because exclusion group. ethnic to more than one racial or group or inclusion of the small of citizens Department counts towards BVAP those who provided responses their who multi-racial on black, white, self-identify as or black and but directly impacts on whether census forms self-identify black excludes those who as plan, Georgia's proposed in terms of group combination with racial or ethnic the number of districts it creates with a ma- other than white. includes all of BVAP, retrogressive black, jority numerically or is self-identify those who whether ex- numerically that Geor- ameliorative. I find clusively any or in combination with other calculating superior gia's BVAPis group, figures. method for racial or ethnic in its BVAP testimony Department’s. Georgia presented to the unrefuted 30, 44, (S.D.Miss.2002); County, No. 01- U.S. 106 S.Ct. 92 L.Ed.2d 25 Colleton (1986)). 3581-10, slip op. at 96. minority voting age adopted have Courts The fact that a district has certain the relevant measure of elec-

population as percentage voting-age of African-Ameri- practical and nor- strength toral both guarantee cans does not that the same majority of reasons. The vast ex- mative percentage of voters will be African-Amer- case, their statistical perts, as this base percentage registered ican. The of Voting analyses voting age population. may who are African-American higher readily determined from age population is BVAP, or lower than although that data, over although census variable variation generally percent- within a few time, registra- is not as mutable as voter age points, among turnout black tion data. See ante at 79. registered white voters varies from elec- districts, state, drawing A its controls tion to election. In ten of the twelve age percentage of minorities majori- benchmark districts with a BVAP But population placed in that district. ty, including Senate 2 and Districts minorities, state, control, not the as a those percentage of registered voters who are a func matter of individual choice and as higher black is than percentage. the BVAP political organization through regis tion of Pl.Ex. IE. implies voting-age See This that like, percentage tration drives and the throughout African-Americans the state registered minority voters.73 Section 5 mobilized, are registration because their prevents any minority’s diminution rates slightly regis- excess of white opportunity representatives to elect of its majority- tration rates ten of the twelve “by choice caused the State’s actions.” minority districts. Vera, Bush v. 116 S.Ct. U.S. concepts percentage Courts’ terms, By phrase its “effective minority voting age population necessary implies franchise”

exercise of electoral comprise majority, to an “effective” any Georgia, it is not the role of other cf. Ketchum, 740 F.2d at 1402 n. one state, minority to create districts that vot choice, minority can elect candidates seg ers can win even when substantial early have varied over time. Until the register fail to voting-age ments of adults 1980s, suggested wisdom conventional or, registered, stay home on elec having super-majorities African-American of 65% day. “Accounting reg tion for lower voter majorities. to were needed create effective among istration and turnout rates Elections, v. Rybicki See State Bd. determining citizens when what constitutes (N.D.Ill.1982) F.Supp. 1114 n. 87 ‘equal opportunity participate an (“The figure general guideline is a 65% process’ creating electoral ‘safe black Department which has been used compensate districts’ to for those rates Justice, reapportionment experts and the amounts to an incentive for and institution minority popu courts as a measure of the apathy.” alization of black voter Johnson (S.D.Ga. Miller, lation in a district needed for F.Supp. v.

1994) meaningful opportunity (quoting Thornburg Gingles, v. 478 voters Dep. Dougherty County Testimony from witnesses resident in the Shinhoster at 36. regis- contested Senate districts indicates that targeted registration campaign has been for a control of tration is a variable well-within the percentage that aims to increase the of Afri- community. the African-American Commu- percent. voters two to three can-American register nity to leaders in Savannah were able 98, Dep. at See White 5,000 African-American voters. See new

122 choice.”); far in the handful cases decided thus of candidate of their elect a cf. cycle, trend indicates 164, redistricting 2000 the UJO, (noting at 97 S.Ct. 996 430 U.S. courts, factual circum- under certain conclusion Department’s Justice stances, willing accept now population majority that a non-white minority retain the proposition that voters 65%, non- in order to achieve a vicinity of their electoral franchise ability to exercise voters, not majority eligible of was white bare, effectively in districts with 50% unreasonable). twenty intervening In the Council, County majority. See Colleton strong indications that years, there are 88, 96; 01-3581-10, slip op. at see also No. virtually in race relations progress Bartels, F.Supp.2d 144 358 Page v. minority rationale for 65% eliminated the (D.N.J.2001). court, applying the One super-majorities. presciently noted guideline,

65% measure, Supreme As a related and elec- changes sociological “emerging considered, as a de stan- Court has facto minority groups characteristics of toral franchise, the for effective electoral dard political attitudes changes and broad minority number of seats voters control. eliminate, alter, substantially may “If our standard using control seats as figure, for a corrective. The 65% need theory very does not reflect a nuanced regular- рarticular, should be reconsidered political participation, it at least has the information and new sta- ly to reflect new superficial advantage appealing to the Ketchum, at data.” 740 F.2d tistical political easily ‘most measured indicia ” Holder, 899, 114 power.’ 512 U.S. at 1990s, began recog courts During Bandemer, Davis v. (quoting S.Ct. 2581 voting age nize that districts with 55% 106 92 478 U.S. S.Ct. majorities preserved opportuni effective (1986) (O’Connor, J., concur- L.Ed.2d 85 voters, minority ties for see v. Clin Jeffers plan, the benchmark minor- ring)). Under (E.D.Ark. ton, F.Supp. ity voters have elected eleven candidates 1990), voters, minority in certain cir while African-American, choice, all from the cumstances, they retained effec believed majority twelve districts with BVAP. Black at even lower levels. voting power tive District 12 benchmark Senate Vera, See Bush v. U.S. unable to elect their candidate been (in redistricting process, S.Ct. measure, Using pro- of choice. this community insisted that ‘safe’ “[t]he [a] posed plan retrogressive would not if be that had a total black district be drawn African-American voters retained abil- 50%”) (internal population of at least ity to control eleven or more seats in the omitted); DeGrandy, quotations state Senate.74 (African-American F.Supp. at 1088 n. 5 majority-minority intervenors contended that district with dis- The number provided proposed plan less than 50% BVAP nonetheless tricts created and the opportunity minority African-Americans with an to number of candidates of choice choice). reasonably expected to elect candidates of Based on the who can be incumbent; popular question defeat a 74. The' fact that black voters in Senate Dis- may not be to defeat the current required trict 12 able what BVAP is to insure that an of retro- white incumbent is not indication opportunity minority population equal has an gression, represents only a because it continu- to elect a candidate of choice in an quo incumbency. ation of the status Council, No, open County —white election.” Colleton three-judge recently As district court another 01-3581-10, op. (emphasis orig- slip at 101 observed, ques- "We must remember that the inal). necessary tion is what BVAPwould be purpose Voting Rights only the of the Act does not end from those districts elected pro- where, here, waning or not the as racism is and mi- into whether inquiry the § 5. It is also plan satisfies posed nority necessary candidates can draw the Senate to consider in most situations necessary support. of white See at 50-52. level infra white reasonably anticipated whether § 2 litigation Voting under of the the numerical vote will enable crossover Act, to Rights Supreme Court refers voters, in terms of strength racially polarized voting voting patterns as BVAP, into effective electoral to translate majori- produce which elections which a ability of the to elect strength, terms majority of African- ty of whites and a At the bot- of choice. minority candidates support Americans each a different candi- line, African-American voters tom effec- polarization legally This racial is dates. franchise tively exercise their electoral § in a 2 case if “a white significant votes, taking into account the when their normally vote ... will defeat the com- bloc crossover, likely white magnitude minority plus strength support bined choice to their candidates of will enable Gingles, ‘crossover’ votes.” 478 U.S. white win, victory is regardless of whether follow, It should at 106 S.Ct. 2752. single a vote. by a landslide or coin, § opposite of this that in a side Voting Versus Crossover White context, racially polarized voting legally in the Pro- Racial Polarization minority if practically insignificant posed Districts likely candidates of choice will be able to number” crossover votes to percent “magic is not a attract sufficient white Fifty Where, here, strength. electoral defining effective win. join minority to which voters equal The extent creates a number of districts plan races to elect a forces with voters of other greater plan than the benchmark key varia- mutual candidate of choice is sufficiently large likely which it is Majority-minority districts are a nec- cohort, ble. and cohesive African-American see remedy in circumstances where essary 50-51, 106 S.Ct. will combine id. at minority support refuse to white voters of white crossover with sufficient numbers candidates, “in communities in but those majori- voting to create an absolute voters minority citizens are able to form which plan retrogressive is not violation ty, the racial and voters from other coalitions with § of 5. are not groups minority ... voters ethnic patterns are not agree I that “[i]f haul, pull, obligation

immune from the polarized voting,” by racially marked —in ground.” De and trade to find common words, if sufficient white other there is Grandy, 512 at 114 S.Ct. 2647. U.S. percentage crossover—reductions requires minority redistricting A no ef “may have little or white support to draw candidates ability preferred to elect can fect on their purposes voters is not offensive to didates” and would evidence absence Act, indeed has a Voting Rights retrogression. Ante at 78. backsliding or in a slighted is not to be virtue “which However, testimony from they invoke the waning meant to hasten statute witness, Eng- Department’s expert Id. Howev- politics.” racism in American strom, candidates that African-American er, redistrieting plan that lowers BVAP necessary attract will unable to minority can- majorities point to the where votes to win of white crossover at least level likely of choice will need didates Eng- See id. proposed districts. support prevail effectuates some white 124 patterns of the Senate district as a presented voting data from Senate

strom has races, elections, whole, possibly municipal rather than a discrete and statewide county Although Eng- unrepresentative part, largest elections. as its such and/or well-accepted statistical study strom’s uses city populous county. or most Additional methods, a large gap there is between county ly, municipal and boundaries do not voting in different levels of white crossover necessarily overlap with the boundaries of that he is unable to types of elections district. See ante at 85-86. considering without which explain. Even predict The use of local election results to probative, type of election is most level of white crossover in future Sen degree mere fact that “the of racial bloc highly questionable, ate races is therefore voting widely varied from election to elec former differ from the latter because the argues against finding legally a tion” election, only type in but also racially polarized voting exists. significant in incomplete, involve an some cases (6th Ohio, 377, Mallory v. 173 F.3d 382 extraneous, group of voters. See Clark v. Cir.1999). (5th County, 21 Calhoun F.3d 96 Cir. 1994) municipal (rejеcting the use of elec provide proba- elections the most predictor equal opportunity tions as a results, involving “elections tive because elect candidates of choice to coun particular office at issue will be more office). ty-wide involving relevant than elections other of- fices,” Lee, Magnolia Bar Ass’n v. in The statewide election results the dis- (5th Cir.1993), F.2d but puted a much higher Senate districts show sample severely data is limited.75 support rate of white crossover for black candidates than in case local elec- majority Engstrom’s finds that tions those districts.76 When white vot- elections, analysis of local which indicates 2, 12, ers Senate Districts and 26 voted sup- percentage small of whites will elections, candidate, in statewide between 23.8% and port probative a black is more 58% cast their ballots for black candidate analysis than his of the district-wide re- over a white candidate. crossover turns in statewide elections. See ante at White Although averaged neither the statewide nor statewide elections 40.3% office, local races are for the relevant the Senate District 41.8% Senate District results the state-level races reflect the and 43.3% in At Senate District 26. aside, Engstrom provided no data from Senate As an I note that the differential multiple 46, ante, and the elections in Sen- percentages provided in footnote for single ate Districts and 12 each involved a elections, both statewide and local which candidate, making African-American it im- 63%, range from are at a level that 28% possible to discern whether the crossover disprove racially tends to the existence of rates in those elections were race-based Cincinnati, polarized voting. See Clarke v. response specific to those candidates. (6th Cir.1994) J., (Boggs, F.3d con- (“One curring) polarized excellent measure of theory “preference 76. Plaintiff relies on a voting or white bloc is the difference argue voting patterns differentials” to that the percentage between the in statewide and local are not dis- of whites who vote elections Judge similar. See ante at 87-88. Like Sulli- given percentage for a candidate and the van, persuaded by theory, I am not blacks who vote for the same candidate. tangential weighing it find to our task in classically polarized races in most of the evidence, by determining here the relative cases, voting-rights figure southern this value of statewide versus local elections as a percent tended to be 80 or more almost all predictor voting patterns pro- of racial candidates.”). posed Senate districts. precincts analy- that African- number of available for trial, conceded Engstrom this level who receive Any explanations American candidates sis.79 of these are more support “good have a crossover of white plausible theory than a that white voters at election. See ante winning chance” of support who refuse for racist reasons to rele- Department challenges the 87. candidates for local office would elec- votes in statewide vance of crossover nonetheless vote to elect black candidates elections, in districts to Senate tions those powerful prestigious to more state- voting in that white crossover Sen- arguing positions.80 wide voting patterns akin to ate races is more relevant, they To the limited extent local elections. seen in disprove local election tend to results elec- Engstrom’s analysis state-level Department’s contention that levels of dis- in each of the contested tion results been, voting white crossover and will tricts, anal- re-aggregation his particularly be, so low as defeat qualified continue in each of the ysis indicating the returns minority in candidates the contested Sen- run they elections as if had been statewide Savannah, largest city ate districts. In districts,77 highly proba- is in District African-Americans Senate sufficiently high white tive evidence minority registered were a voters dur- candidates of voting crossover to enable cycles. the 1995 and 1999 ing election See prevail. The lower rates of white choice to Dep. Shinhoster at 36. White crossover may specific in local crossover elections rates, Engstrom, no according were district, of the pockets to those Nonetheless, greater than 10.6%. African- office,78 may be specific to the may be Americans won two of the three runoff statistically accurate due to smaller less analysis was elected in Engstrom's re-aggregation indi- At the time Adams 77. elections, twenty-one that out of cated Savannah’s BVAP was less than 52% eighteen candidate won African-American majority registered whites were a voters. outright, pluralities in received substantial Dep. at 36. This under- See Shinhoster candidates, multiple primary two races with Department’s statistical evidence in cuts the top to advance to the runoff for the sufficient possible ways: in fact re- two either Adams single vote-getters, election. two and lost vote and ceived of the white crossover 20% at 11-12. See Def.Ex. 611 statistically Engstrom's local election data is elec- questionable, or Adams was able to win that, rule, general the lowest I note as a city where black voters were a nu- tion in a rates are seen in elections to local crossover office, mayoral races or elections cross- executive merical with minimal white county to the board of commissioners. over. Floyd Engstrom’s data indicates comparison 80.A white cross- 16.6% Adams, the incumbent African-American attorney over rate in the district election Savannah, mayor received 2.8% County, white crossover Bibb and the 46.9% votes in the election and of white white 8.7% Attorney for state General rate in election he won in 1995. votes in the runoff when may It be that is a co- is illustrative. there testimony Department provided has also who will not hort of white voters community politicians and leaders from local county Savannah, support candidate district a black who testified that Adams in fact roughly attorney simply they are racists. support of white because received 20% However, election. See P. Jackson implausible voters in his 1995 it to assume that is (“[W]e crossover Dep. at 29 calculated white significant portion white racists of those neighborhood of 20 to be somewhere in the support a black can- would turn around and 36; Dep. percent.”); Johnson at Shinhost- H. Attorney state General. didate for Dep. er at 41. candidates competed ante at 62. African-American they where head-to- elections candidates.81 against head white three of the four Macon also won least city elections.85 In Africаn-American won both elections to state candidates victories, are not isolated to be These re- Engstrom’s expert House included aberrations, 45; ante at dismissed as see n. Dougherty County port.82 the two that the relative lack of they demonstrate races, candidate one African-American Engstrom dis- white crossover won, is no and the other lost.83 There legally prac- cerns local elections racial that reveals record evidence To the extent these tically insignificant. Dougherty makeup of the House districts. consideration, worthy of local elections are in County majority white but was posi- they support Georgia’s tend to basic African-American residents percentage a fair tion: that African-Americans have 48; increasing. Dep. See Sherrod at is of their opportunity to elect candidates 16; Dep. D. at 66. Dep. at Williams White in where the black choice even districts in the record to indi- There is no evidence voting age population registered voters African-American candi- cate whether the fifty percent numbers or less. Albany mayoral races won or dates in the lost. Opportunity 3. Fair to Elect Minori- County, county In dominant Bibb ty Candidates of Choice three African- Senate District at least prevent § my It is view that 5 does not county-wide elec- Americans won adopting redistricting plan, a state from County majority-white, tions.84 Bibb leg- of African-American blessing with the approximately 43% BVAP and black islators, roughly “packed” voter 40%. See that reduces concentra- registration soft, relatively Floyd mayor's was at 81. Adams won the race in American voters 72.2%. contrast, Taylor supported by Edna Jackson lost the election for an was 1995. 96% at-large city council seat in but won in black voters. See Def.Ex. 611 at Dep. Interestingly, 1999. See E. Jackson at 8. Engstrom’s report indicates that she received witnesses, Department's 84. Two of percentage an identical of white votes in both Barnes, currently Abrams and serve on the races, meaning victory that her in 1999 was County Bibb Board of Education and defeated relatively higher likely result of African- Engstrom's white candidates to win election. turnout, depend and did not on a candidate, data indicates that third black voting preferences. shift in the white See Hutchings, white vote received 34.2% (Savannah Def.Ex. 611 at 3-4 Council At- in his 1994 election. African-Americans Runoffs). Large, 1999 therefore won two out of four elections ana- lyzed Engstrom's report, and won an addi- countywide testified that White he won his 1994 elec- tional election not included in the tion to the state House with more than data set. 70% Dep. judicial the vote. White at 63. I take Roberts, notice that Lawrence R. the African- Ellis, mayor, 85. Jack Macon’s incumbent won American candidate in House District primary by points, percentage the 1999 six has served in the State House since 1992. receiving of the white crossover voter. 10.4% <http://www. legis.state.ga.us/Legis/1995— See ¶ (Ellis 2). judicial Decl. I take that notice (visited > 96lhouselgahml62.htm March City Brenda Youmas was elected to the Coun- Timley prevailed cil in 1995 and that James 2002). <http://www.cityofma- his 1999 council race. (visit- Taylor con.net/CityDept/council/members.htm> 83. Jane the coroner’s race and won 22, 2002). Wright county ed lost election to chair the March There is no evidence to 18; Dep. Wright commission. See White indicate whether the African-American candi- Dep. Wright’s support among at 14. African- date in the fourth Macon race won or lost. process,” Rybicki, out of the electoral long preserves so as it tions of black voters for minorities to opportunities or fair equal (cid:127)F.Supp. (dissenting opinion), at 1189 may It well of choice.86 elect candidates longer edge super-ma- no need the those of black voters super-majorities jorities provide, opportu- diminishes their plan create “robust” under benchmark nity to influence elections elsewhere and to elect a candidate of choice. opportunities carry “threatens to us further from the the law of unintended conse But under goal political system of a which race no conditions they may also create quences, De longer Grnndy, matters.” 512 U.S. at “unreasonable,” “unfair,” “un that are J., (Kennedy, S.Ct. concur- minority voters those equal,” to both ring). proposed plan provides A a fair “wasted,” to the whose votes districts opportunity greater to elect the same or unnecessary find it point they may number of candidates of choice than the vote, non-minority turn and to out provides § benchmark is entitled to in those districts whose preclearance. by the might “submerged” terests well be turn point they to the that super-majority Weight 4. Relative Accorded to Ex- process. away political from the Gin Cf. 68, Lay Testimony pert 478 at 106 S.Ct. 2752. and

gles, U.S. countenance, Act does not let Voting Right responsibility, It is our as the triers of a require, alone such result. fact, relevance, to determine the credibili- Voting Rights The Constitution and the pre- of ty, proper weight the evidence minority guarantee victory Act do not sented. This case need not be decided candidates, A only equal opportunity. but solely expert testimony. on the basis of minority super-ma of state’s maintenance demonstrated, amply As this trial has sta- jorities particular a district is re within science, tistics is an inexact made more so by § quired by 5 when necessitated “inherently uncertain” nature of polarized voting, legally significant racially Cromartie, data, voting Easley v. behavior ineligible minority of vot large numbers 247, 1452, 121 149 582 U.S. S.Ct. ers, to the effective exer or other barriers (2001), already-outdated L.Ed.2d 430 of franchise that are outside cise electoral Abrams, data, 521 at 100- census see U.S. minority group. the control of the There Supreme 117 S.Ct. 1925. The Court minority group of a to a right is “no vested context, noted, previously in a related majority particular magnitude a unrelat that in another election results oppor of a reasonable provision ed to the “underscore the weakness of the Justice representative.” Ket tunity to elect a methodology calculating Department’s chum, Moreover, the 740 F.2d at 1418. black-preferred of a candi- the likelihood super-majorities, even continuation of per- on strict racial winning date based progress has been made sufficient so when 117 1925. centages.”87 “fenced Id. S.Ct. minority longer voters are no Abrams, Indeed, Department the Justice had Georgia had maintained the if heavy predicted minority concentrations of African-American candidate of choice that a certain of its Senate and House Georgia’s voters in prevail would not be able to districts, metropolitan particularly At- district, Congressional which had a Fourth area, may lanta black voters in those districts Cynthia McKinney, who is BVAP of 33%. cognizable § a 2 claim based on have had Afriсan-American, nonetheless won the elec- through packing. See dilution of their votes Quilter, tion. U.S. at 113 S.Ct. 1149. Washington); Ry Congressman Harold obligation accept sta have no Courts bicki, F.Supp. (crediting at 1114-1115 Magn as conclusive. See tistical evidence (“[T]he that the inclu Ass’n, testimony of black aldermen 994 F.2d at 1149 olia Bar in their wards neighborhoods sion of white any authority, offered plaintiffs jeopardize their re-election none, would can find for their assertion and we chances); Rockefeller, v. Wright may rely only on the district court cf. 603, 11 L.Ed.2d 512 84 S.Ct. determining whether U.S. expert conclusions (1964) (intervention Congressman Adam significant.”). legally bloc white Powell).88 Clayton I therefore accord statistical evi does not mean that This hand, support Brown’s par great weight Senator rejected out dence should be district, for his own as well Senator and uncontra- ticularly weighty when it is probably will experts or Thomas’ concession she testimony from other dicted in her district.89 strong win re-election lay “In the face of witnesses. Lewis, Brown, I broadly, consider general ... statements More statistical case un polls carry to be witnesses with played no role at the Walker that race politics and knowledge Georgia matched Teague County, v. Attala weight.” little Cir.1996). (5th voting strength, with the African-American F.3d While being Brown and Senator Walker qua statistics the record are Senator statistics minority expert opinions especially attuned to the level meaningful, flawed, voting strength necessary contradictory, highly varied choice seat. lay testimony candidates of to win Senate consistency their statistics; Epstein’s I as reinforce and the African- consider statistics and their own political in of assessment of the legislators who testified favor ment this in ground Georgia, much situation on the rather proposed plan of the rendered more support plaintiffs main for the opinions. than the probative argument. the con- Review of this record confirms testimony I of testimony of other courts that the have also considered clusions leaders, witnesses, lay I minority political particularly Department’s although of districts, in to pales importance their believe it the testi- regarding incumbents own Lewis, Brown, mony of and and highly probative is evidence of the minori- Walker percentages necessary expert to ensure contin- witnesses. To the extent ty minority discrepancies in candidates exist between declarations electing ued success See, County depositions Department’s lay and of the e.g., of choice. Colleton Coun- (testi- 01-3581-10, witnesses, cil, accept I the latter as more cred- slip op. at 110 No. ible, mony Congressman Clyburn); represents because it the witnesses’ of James words, Ketchum, (testimony adoption 740 F.2d at 1415 of own rather than the Rybicki speculated retrogression, 88. The Court that a three- District 2 is based less on and Supreme judge district court and the Court garden-variety political more on concerns: redistricting may upheld plan that plan impinges political proposed on her heavily divided Manhattan into three white County bringing in a third turf in Chatham Congressional heavily districts and one mi- senator, likely and makes it that she will have Powell, nority Congressional district because campaign harder than in her- most recent incumbent, the African-American intervened election, when she received more than 77% support a defendant in legitimate objections the vote. Her are and plan. Rybicki, F.Supp. at 1118 n. 97. See political standpoint, understandable from a implicate Voting Rights Act. but do not testimony, 89. Based on Senator Thomas' I opposition to Senate infer that her jeopardize strug- that would what he has partially prepared at least statements It in- gled inherently so hard to win. is testimony is deposition the counsel. Brown, credible that Senator Senator comprehensive, permits and also more Walker, all but one of the African- explain and elaborate on witnesses of the Georgia American members Senate in the declarations. contained statements place this court to their would invite Sen- Georgia’s Purpose Plan risk, B. at serious cause ate seats who, all, a-majority after are voters — majority that the Sen- agree I equal oppor- their constituents —to an lose in this redistrict- purpose advancing ate’s tunity to elect candidates of their ante at non-retrogressive. is See ing plan judicially changes choice. The noticeable pur- legislators had a dual Georgia’s political landscape the South in existing minority voting maintain pose: to Georgia particular, general, avoid the “waste” of black strength but days “Massive Resistance” to votes, major- maintain a Democratic and to present, expressed corroborate the confi- Assembly. houses of the State ity both Georgia’s dence of African-American legis- Dep. at 20-21. The role that Meggers See steady lative leaders that the rise in white played legislators African-American voter crossover from zero to substantial House, redistricting plans for the drafting will continue. numbers This confirms to and their near-unan- Congress, proposed redistricting me that the of Sen- plans, support imous for those 2, 12, ate Districts and 26 would not cause a reliable indication that has no retrogression ability of African- retrogressive purpose. American voters in those districts and strengthen position The desire statewide to exercise their franchise effec- political party one relative to the other is tively. purpose, although a retrogressive not I majority opinion have read the here, circumstances, present there open concurrence with care and an mind. might retrogressive it have a effect. where However, I persuaded by Judge am not respect redistricting, With interests comprehensive Sullivan’s and well-written party Georgia’s pri- and of the Democratic majority opinion by the concurrence. largely in marily Democratic voters are points four I consider to Neither addresses Although tandem. that does not immunize be crucial to the resolution this case. proposed redistricting plans from com- First, a three-judge a court is trier of 5,§ plying argues against finding it a jury, judicial with a triers of fact fact. As purpose. discriminatory may reach different conclusions about evidence, III. Conclusion in- probative value of items of expert testimony. cluding Lewis, testimony I find that Second, Brown, persuasive response no provides and Walker the State there is my that that the number of preponderance with a of evidence out- observations districts, majority-minority measured weighs expert testimony and the testi- I courts to have testified for and most mony lay of the witnesses who BVAP—which highly find to be a and intervenors. Con- considered the issue Department life, minority voting probative gauge and gressman Lewis has devoted his once, by one from the strength it more than to advance the risked —increases plan, and the number of minor- voting rights. of African-American benchmark cause likely to elected redistricting plan ity choice candidates be He would not advocate a Act No. reapportionment plan, gressional proposed plan is the same under the defendants; 2EX11, and it is against plan. under the benchmark Third, any legal I have not discovered AD- FURTHER ORDERED majority propo- support authority shall enter final that the Clerk JUDGED plan pre- § that a requires sition re- of defendants with judgment favor (whatever means), a “robust” serve reappor- spect Georgia’s State minority probability that pre-existing 1EX6, against plan, Act No. tionment majori- prevail, and the candidate of choice plaintiff. have held that ty none. Other courts cites THE IT FOR IS SO ORDERED preserves or increases THREE-JUDGE COURT. districts where number of opportunity to “equal” have a “fair” or are not their candidates of choice

elect at 43^44.

retrogressive. supra See

Fourth, that a trial court is the notion counsel, which arguments

bound principal- case as

may perceived this polarization, racial will

ly concerned with Plaintiff, Gayle Harper BRUNK, surprise to most trial great come as a jurors that the judges. court We instruct v. of the law- arguments “[statements PLC, Defendant. BRITISH AIRWAYS Although it yers ... are not evidence.”90 attention where may be easier to focus our CIV.A.00-00764(HHK). No. it, expert on the wit- lawyers direct Court, United States trial, the volu- who testified live nesses District of Columbia. testimony from other wit- minous written part evidentiary equally nesses is April responsibility in this case. record Our it assign evidence in toto and to review the it weight deserves.

Therefore, respectfully I dissent from majority respect decision with to the redistricting plan.

JUDGMENT Rule of Proce-

Pursuant to Federal Civil

dure 58 and for the reasons stated Opinion

court in its and Order docketed day, hereby

this same it is ADJUDGED that the

ORDERED and in favor of judgment

Clerk shall enter final

plaintiff respect Georgia’s State plan, Act No. reapportionment

House

2EX23, Georgia’s United States Con- Jury District Columbia at 2-5. Civil Instructions Standardized The court notes voting patterns to demonstrate appears information to be crucial to the meth- existing, calculated crossover in the odology employed by Epstein, pro- it was not Tr., Day proposed, districts. not the or vided defendants intervenor-defendants 1, p.m. at 71. shortly beginning of the trial. until before votes crossover enough white Democratic voting pat- level analyze precinct the district overall. terns, percentage to win and to estimate to vote for “crossed over” voters who white of Afri- the races also considers Epstein in each of candidates African American incumbents, that a but finds can American Districts. U.S.Ex. existing Senate inappropriate for these analysis is probit from on are based data These estimates an races with the recent 200 races. Of general elections. three 1998 statewide incumbent, only once African American white crossover average estimated Id. The incumbent de- African American was the districts ranges voting those 1992; lost Hildred Shumake feated. that Epstein Id. testified 24.73%to 57.39%. election, however, Shu- at the time of range to be statis- consider this he did not indictment on extortion make was under es- because the crossover tically significant from the State suspended charges distributed, all regularly timates were incumbents African American Senate. expected fell within the two but “outliers” as 35% as low have districts with run that, af- further testified range. Epstein Therefore, to assess and have won. BVAP his expanded he deposition, his first ter in- American of an African the likelihood crossover analysis to look at black candidate who losing to a white cumbent indicators levels as registration and black choice, Epstein a candidate of was not 2/4/02, Tr., mobilization. ‍​‌​‌‌‌​‌​​​​‌‌​​‌‌​​‌‌​‌​​‌‌‌‌​‌‌‌‌‌​‌‌‌‌‌​‌‌‌‌​‍voter the Afri- open seat data with combined analyses additional p.m. at 68-69. These This elections. can American incumbent the court. part of the record before are not opportunity point of yielded equal an redirect, However, Epstein testified BVAP. 37.4% when “exact same result” got that he analyses, finding he ran these an African American point at which levels registration and black crossover greater chance candidate has 50% distribution,” “very “tight were in a incumbent occurs unseating a white Tr., 2/5/02, a.m. at 98. few outliers.” cur- District 12 is BVAP. Senate 56.50% regarding the reach conclusions To his Meyer Van rently represented Senator non-retrogressive effect of the Burén, existing man. district a white upon analysis plans, Epstein only relies an BVAP, be re- which would has a 55.25% in- open The database seat elections. pro- to 50.66% BVAP under duced Ep- open-seat cludes elections. that, Epstein testified ac- posed plan. point equal opportu- stein finds calculations, African to his an cording BVAP, which nity is 44.3% means candidate of choice would a candi- a 50-50 chance of electing “there’s being elected in a 50% chance of less than open in a district with an date of choice” and a white with 55.25% district BVAP Tr., 2/5/02, and with 44.3% BVAP. seat incumbent; probability of the estimated 29; ex- Epstein a.m. at Pl.Ex. 25 at choice American candidate of an African by noting that over 90% plains this result in- in a with white being district elected support Demo- of African American approxi- and 50.66 BVAP falls cumbent crats, sup- over 60% of white voters while mately 29%-30%. Thus, Id. at n. 14. port Republicans. trial, pro- day Epstein On the first supported candidate an African American curve,” previously had duced an “S which in a primary to win a usually would able

Notes

notes U.S. States the United test simple doctrinal no “there ing removed was Army Air Field Hunter ra significant legally existence The inclusion for the proposed district. that Sec Gingles noted voting,” dis- permitted the cial bloc army base would necessarily fact-sensi population, trict, inquiries in need of tion was which 2752. Justice avoiding a reduction 106 S.Ct. tive. Id. population while add factors 1-14 192: a number suggested U.S.Ex. Brennen in BVAP. See ¶¶ Section 214-15. under testimony); may USPFF consider that courts (Meggers’ normally will vote State, white bloc Moreover, despite its [A] minori- strength the combined not forced contrary, was defeat to the suggestion votes plus white “cross-over” support complying ty between to choose significant legally Voting to the level rises Clause Equal Protections of white The amount voting. that the dictates bloc is true white Act. It Rights “minimize generally can at times force may vote bloc person-one one to elect ability voters’ to reduce cancel” black redistricting state

notes Ms. Thomas Repub- to vote urged people ence, always opponents her white against to vote Party but primary, lican precincts. U.S.Ex. majority won white Party primary. Democratic White (Thomas). 704, Other witnesses 104:11-19 the flier as “a call Id. White characterized high County testified Chatham rally together arms for voters white polarized voting, i.e. levels a black candidate.” U.S.Ex. to defeat their for candidates cast ballots One NAACP same race as themselves. ¶ 12. (Alderman Jones); (Senator See, U.S.Ex. 53:20-54:5 e.g., U.S.Ex. 104:11-19 12; ¶ Jackson, Jackson); (Dr. U.S.Ex. Thomas); U.S.Ex. 22:4-24 U.S.Ex. 22:20-25:8, 65:7-14, (County Bd.Comm. NAACP); 27:23-28:15 68:1-11 U.S.Ex. 9; (County ¶ Odell); Bd. NAACP); 66:24-67:3 (Johnson, U.S.Ex. U.S.Ex. U.S.Ex. Rivers). 36:2-7; 709, 35:5-20; (Alderman Comm. 63:17-65:4

Case Details

Case Name: Georgia v. Ashcroft

Court Name: District Court, District of Columbia

Date Published: Apr 5, 2002

Citation: 195 F. Supp. 2d 25

Docket Number: Civ.A. 01-2111(EGS)H

Court Abbreviation: D.D.C.