United States v. Johnson Jurisdictional Statement

    No.

3tt %  Btxpttm (Emrt xtt %  Mnxtth Btntm
O c t o b e r  T e r m , 1995

U n it e d  S t a t e s  of  A m e r ic a , a p p e l l a n t

D a v id a  Jo h n s o n , e t  a l .

ON APPEAL FROM THE UNITED STATES DISTRICT COURT 
FOR THE SOUTHERN DISTRICT OF GEORGIA

JURISraCTIONAL STATEMENT

Drew S. Days, III 
Solicitor General

Deval L . P atrick 
Assistant Attorney General

P aul Bender 
Deputy Solicitor General

James A. F eldman 
Assistant to the Solicitor General

S teven H. Rosenbaum  
Miriam  R. E isen st bin  

Attorneys
Department of Justice 
Washington, D.C. 20530 
(202) 5U-2217



QUESTIONS PRESENTED

1. Whether the district court’s congressional districting 
plan for the State of Georgia contains a greater deviation 
from the constitutional one-person, one-vote requirement 
than is necessary or justifiable.

2. Whether the district court’s congressional districting 
plan for the State of Georgia deprives African-Americans 
of an equal opportunity to participate in the electoral 
process and elect candidates of their choice, and thereby 
violates Section 2 of the Voting Rights Act.

3. Whether the district court’s congressional districting 
plan for the State of Georgia departs from state district­
ing policies to a greater extent than necessary to cure the 
constitutional violations in the State’s plan, in contraven­
tion of this Court’s decision in Upham v. Seaman, 456 
U.S. 37 (1982).

(1)



II

PARTIES TO THE PROCEEDING

In addition to the parties named in the caption, the 
plaintiffs below were Pam Burke, Henry Zittrouer, George 
L. DeLoach, George Seaton, James Rhodes, Gary Heard, 
Daniel Morrel, and Donald Shirah. The defendants below 
were Zell Miller, Governor of Georgia, Pierre Howard, 
Lieutenant Governor of Georgia, Thomas Murphy, 
Speaker of the House of Representatives of Georgia, and 
Max Cleland, Secretary of the State of Georgia. Max 
Cleland has been succeeded as Secretary of State by Lewis 
Massey. Intervenor-defendants were Lucious Abrams, Jr., 
Rev. G.L. Avery, William Gary Chambers, Sr., and 
Karen Watson.



TABLE OF CONTENTS
Page

Opinions below........... ..... ..............................................-....  1
Jurisdiction..................... —.......... .......................................  1
Statutory provision involved..... .................................. .....  1
Statement.......... ................ ........ ..... ...................... ............  2
The questions raised are substantial............ ........ ......... . 10
Conclusion.......... ....... ....... ..... .............. .............. ........ . 17
Appendix A _________ _______ _______ __ ______ ___  la
Appendix B ....................... ............ .......................... ...........  30a
Appendix C_____ _______ __ _____ ____ ___ ______  39a
Appendix D ...................................................... ...................  47a

TABLE OF AUTHORITIES
Cases:

Bushee V. Smith, 549 F. Supp. 494 (D.D.C. 1982),
aff’d, 459 U.S. 1166 (1983) ........................ ..........  3, 22

Chapman V. Meier, 420 U.S. 1 (1975).................. . 11,14
Collins V. City of Norfolk, 883 F.2d 1232 (4th Cir.

1989), cert, denied, 498 U.S. 938 (1990)______ 18
Connor V. Finch, 431 U.S. 407 (1977) ________11,14, 26
Houston V. Lafayette County, 56 F.3d 606 (5th Cir.

1995) _______ _____ ______ _______ ________  16
Jeffers V. Clinton, 730 F. Supp. 196 (E.D. Ark.

1989), aff’d mem., 498 U.S. 1019 (1991) ........... . 16
Johnson V. DeGrandy, 114 S. Ct. 2647 (1994).......  23
Johnson V. Miller, 864 F. Supp. 1354 (S.D. Ga.

1994), aff’d and remanded, 115 S. Ct. 2475
(1995) ............ ........... ............ ................ ...3,15,18,19, 21

Karcher V. Daggett, 462 U.S. 725 (1983)___ __ ___11,14
Kirkpatrick  V. Preisler, 394 U.S. 526 (1969)______11,14
Miller V. Johnson, 115 S, Ct. 2475 (1995) .....2, 3, 4,17, 25
Reynolds V. Sims, 377 U.S. 533 (1964) ........,....... . 11
Thornburg V. Gingles, 478 U.S. 30 (1986)..9,12,17, 18, 21

(m)



IV
Cases—Continued: Page

Upham V. Seamon, 456 U.S. 37 (1982)............ 3,10, 23, 25
Wells V. Rockefeller, 394 U.S. 542 (1969)_______ 14
W esberry V. Sanders, 376 U.S. 1 (1964) ......... ........  10,14
White V. Weiser, 412 U.S. 783 (1973) .....................  14, 25

Constitution and statutes:
U.S. Const.:

Art. I ................ ... ........ .............. -..........................  10
§ 2 ....... ................ ...... ................ ....... .........  10

Amend. XIV (Equal Protection Clause)___ 2
Voting Rights Act of 1965, 42 U.S.C. 1971 et seq.:

§ 2,42 U.S.C. 1973............. ............. ............... .-.passim
§ 2 (a), 42 U.S.C. 1973( a ) .......................... .......  2
§ 2(b), 42U.S.C. 1973(b)_________________  2
§ 5, 42 U.S.C. 1973c .................. ...... .................... 3, 4, 24



3n t!|(̂  OIxmrt trf %  ^tato
O c t o b e r  T e r m , 1995

No.
U n it e d  S t a t es  of  A m e r ic a , a p p e l l a n t

V.

D a v ida  Jo h n s o n , e t  a l .

ON APPEAL FROM THE UNITED STATES DISTRICT COURT 
FOR THE SOUTHERN DISTRICT OF GEORGIA

JURISDICTIONAL STATEMENT

OPINIONS BELOW
The remedial order and opinion of the three-judge dis­

trict court (App., infra, la-29a) is not yet reported. The 
three-judge court’s earlier opinion with regard to the Sec­
ond District (App., infra, 30a-38a) is also not yet re­
ported. This Court’s prior opinion in this case is reported 
at 115 S. Ct. 2475. The three-judge court’s original lia­
bility opinion is reported at 864 F. Supp. 1354.

JURISDICTION
The judgment of the three-judge district court was 

entered on December 13, 1995. The United States filed 
a notice of appeal on January 12, 1996 (App., infra, 
47a-48a). The jurisdiction of this Court is invoked under 
28 U.S.C. 1253.

STATUTORY PROVISION INVOLVED
Section 2 of the Voting Rights Act of 1965, 42 U.S.C. 

1973, provides:

( 1)



(a) No voting qualification or prerequisite to vot­
ing or standard, practice, or procedure shall be im­
posed or applied by any State or political subdivision 
in a manner which results in a denial or abridgement 
of the right of any citizen of the United States to vote 
on account of race or color, or in contravention of 
the guarantees set forth in section 1973b(f)(2) of 
this title, as provided in subsection (b) of this section.

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

STATEMENT

In Miller v. Johnson, 115 S. Ct. 2475, 2490 (1995), 
this Court held that the Eleventh District in Georgia’s 
congressional redistricting plan violated the Equal Pro­
tection Clause because “[r]ace was the predominant, over­
riding factor explaining the fGeorgia] General Assembly’s 
decision to attach to the Eleventh District various ap­
pendages containing dense majority-black populations,” 
and because that use of race was not “narrowly tailored 
to achieve a compelling interest.” On remand, the three- 
judge district court held that the Second District in Geor­
gia’s plan—also a majority-minority district—was simi­
larly unconstitutional. After the legislature was unable 
to enact a new plan, the district court devised its own 
remedial plan. This appeal seeks summary reversal of



that plan on the ground that it violates constitutional one- 
person, one-vote requirements, Section 2 of the Voting 
Rights Act, 42 U.S.C. 1973, and the requirement that a 
court-imposed redistricting plan depart as little as possible 
from the State’s redistricting policies, see Upham v. 
Seamon, 456 U.S. 37 (1982).

1. The 1990 census entitled the State of Georgia to 
an additional—eleventh—congressional district. Prior to 
that census, African-Americans, although 27% of Geor­
gia’s population, constituted a majority of the voting age 
population (VAP) in only one of Georgia’s ten congres­
sional districts—the Fifth District, located in the Atlanta 
area. See Bushee v. Smith, 549 F. Supp. 494 (D.D.C. 
1982), aff’d, 459 U.S. 1166 (1983). In 1991 and early 
1992, the Georgia legislature successively enacted two re­
districting plans. Each of them contained two majority 
African-American districts: the Fifth District, in the At­
lanta area, and the Eleventh District, extending eastward 
from the DeKalb County suburbs of Atlanta. The De­
partment of Justice interposed objections to both plans 
under Section 5 of the Voting Rights Act, 42 U.S.C. 
1973c.

In 1992, the state legislature enacted a plan contain­
ing three congressional districts in which African Ameri­
cans constituted a VAP majority. See Miller v. Johnson, 
115 S. Ct. 2475, 2482-2484 (1995). That plan received 
preclearance, but the Eleventh District was found uncon­
stitutional by a three-judge district court. Johnson v. 
Miller, 864 F. Supp. 1354 (S.D. Ga. 1994).

This Court affirmed. The Court held that the district 
court had correctly subjected the Eleventh District to 
strict scrutiny because race had been the “the predomi­
nant, overriding factor” in the drawing of the Eleventh 
District. 115 S. Ct. at 2490. The Court found that the 
State’s plan “subordinated traditional race-neutral district­
ing principles * * * to racial objectives.” Id. at 2488. 
The Court noted that the plan’s extension of the Eleventh 
District to Savannah was occasioned entirely by the desire



to place African-American population in the City of 
Macon (Bibb County) in the Second District, id. at 2489, 
and that many counties and precincts were split for the 
sole purpose of increasing the Eleventh District’s black 
population, ibid.

Applying strict scrutiny, the Court concluded that the 
State’s use of race in drawing the plan was not justified 
by the objective of complying with Section 5 of the Voting 
Rights Act. The Court held that there was “no reason­
able basis to believe that Georgia’s earlier enacted ;[two 
majority-minority district] plans violated § 5,” because 
they were not retrogressive and because ‘ftjhe State’s 
policy of adhering to other districting principles instead 
of creating as many ma j ority-minority districts as possible 
does not support an inference” of discriminatory purpose 
under Section 5. 115 S. Ct. at 2492.

2. On remand, the district court permitted the plain­
tiffs to add parties and amend their complaint to chal­
lenge the Second Congressional District (the other new 
majority-minority district that had been created by the 
State in 1992). App., infra, la-2a. The district court 
refused, however, to allow any Georgia voters to inter­
vene in defense of the Second District. On October 30, 
1995, the district court held a hearing on the constitu­
tionality of the Second District. On December 1, 1995, 
the district court issued an order holding the Second 
District unconstitutional, for essentially the same reasons 
that this Court had held the Eleventh District unconstitu­
tional. App., infra, 30a-38a.

3. After this Court’s remand order, the district court 
offered the legislature an opportunity to formulate a new 
districting plan, but the legislature was unable to do so. 
Each chamber of the state legislature passed a different 
plan and, on September 13, 1995, the legislature notified 
the court that it was unable to redraw the map. App., 
infra, 2a.

4. In October, 1955, the district court issued two or­
ders. The first (R. 325: Order of October 18, 1995) called



upon the parties to- submit remedial plans that corrected 
the faults of the unconstitutional plan while making the 
fewest changes in it. The second order asked the parties 
to file remedial plans “that eomport[] with Georgia’s his­
torical districting principles, Constitutional requirements, 
and the Voting Rights Act.” R. 329; Order of October 
20, 1995. The second order requested the parties not to 
base their proposals on the plan that had been ruled un­
constitutional, but rather upon the first plan Georgia had 
enacted in 1991—a plan containing two majority-minority 
districts.

The plans offered to the district court during the re­
mand proceedings fell into two general categories. In 
one category were four plans submitted by the Abrams 
Intervenors, one submitted by Congressmen John Lewis 
and Newt Gingrich, and one passed by the State House. 
See R. 296, 318, 319, 344, and 347, and Abrams’ Inter­
venors’ Exh. 37. Those plans each contained at least two 
districts with African-American VAP majorities: the 
Fifth, in the Atlanta area, and the Eleventh.'^ In each 
plan, the Eleventh District had between 50%-53% black 
VAP. In all but one, the Eleventh District was located 
in the same east-central part of Georgia as the Eleventh 
District in the State’s three 1991 and 1992 plans, but 
with counties reunited and the extension to Savannah 
removed. Of this group of plans, the Abrams Plan A 
(R. 296 and 319) had the lowest overall population 
deviation® (0.29), while the plan passed by the state 
House (Abrams’ Intervenors’ Exh. 37) had the highest 
overall deviation (1.92).

One of the plans, which was known as the Abrams Plan A 
(R. 296 and 319), had three African-American majority VAP 
districts.

2 We use the term “overall population deviation” to mean the 
difference, expressed as a percentage of the population of an ideally 
apportioned district, between the populations of the district with the 
largest population and the district with the smallest population.



In the second category were the plaintiffs’ proposed 
plans and the plan that had been passed by the State 
Senate. R. 342, 367, 258, and 295. In each of them, only 
the Fifth District in the Atlanta area had an African- 
American VAP majority. None of these plans contained 
another congressional district in which the African- 
American VAP was greater than 40%. The range of 
overall population deviations for these plans was from 
0.52 (for Remedy 4X and Remedy 4XR, see R. 367, 
Exhs. B and D) to 1.78 (for Remedy 4, see R. 342). 
After a hearing on October 31, the district court ordered 
plaintiffs’ counsel to revise one of the plaintiffs’ plans 
(Remedy 4X) in order to reduce the overall population 
deviation and to submit that plan for comment. The 
court called upon the parties to submit their comments 
by November 22. R. 363.

On November 22, the United States submitted a brief 
(R. 348), acompanied by an “Illustrative Plan,” arguing 
that the plaintiffs’ proposed Remedy 4X did not meet the 
requisite standards of population equality, made more 
changes than necessary to cure the constitutional violation, 
and violated the Voting Rights Act by including only one 
majority-minority district. The Illustrative Plan was in­
tended to illustrate that a plan could be drawn with two 
compact majority-minority districts and minimal popula­
tion deviations. It splits only two counties outside the 
metropolitan Atlanta area. Its overall population devia­
tion of 0.19% was the lowest overall population devia­
tion of any plan before the district court. The plan in­
cludes two compact majority African-American districts— 
the Fifth and the Eleventh. A map of the Illustrative 
Plan and demographic data for the district appear at 
App., infra, 44a-45a.®

® The maps of the district court’s plan, App., infra, 42a, the 
Illustrative Plan, App., infra, 44a, and the first plan the state 
adopted after the 1990 census, App., infra, 46a, were reproduced 
on computers at the Department of Justice so that the colors and 
legends on the maps would be uniform.



5. On December 13, 1995, the three-judge district 
court adopted the plan from which this appeal is taken. 
That plan, a map of which appears at App., infra, 41a, 
closely resembles the plan (Remedy 4XR) that the plain­
tiffs submitted in response to the court’s order following 
the October 31 hearing. It has an overall population de­
viation of 0.35%. It includes only one majority-minority 
district (the Fifth), which contains a 57% African- 
American VAP. App., infra, 6a, 25a. See also App., 
infra, 43a (demographic data for court’s plan). The 
court ordered the State to adhere to that plan “until the 
legislature convenes to change it * * * following the 
millenium census, or before.” App., infra, 28a-29a.

a. The district court recognized that, “[i]n fashioning 
a remedy in redistricting cases, courts are generally lim­
ited to correcting only those unconstitutional aspects of 
a state’s plan,” because such a “minimum change plan acts 
as a surrogate for the intent of the state’s legislative body.” 
App., infra, 2a-3a (citing Upham v. Seamon, 456 U.S. 
37 (1982)). The court held, however, that “Georgia’s 
current plan was tainted by unconstitutional [Department 
of Justice] interference,” App., infra, 4a, and that the 
court therefore “c[ould] not use Georgia’s current plan as 
a surrogate for the legislature’s reapportionment policies 
and goals.” App., infra, 5a. The court also stated that 
“any remedy even minimally disruptive to the current plan 
would necesarily have resulted in drastic changes” because 
the two districts that had been declared unconstitutional 
were “on opposite sides of the state,” and “it is a virtual 
impossibility to change less than we have.” App., infra, 
6a.

With respect to the goal of population equality, the 
court noted that “federal courts are held to stricter stand­
ards than legislatures,” and that the court therefore was 
“particularly constrained to create a remedy with the 
lowest population deviation practicable.” App., infra, 7a. 
The court stated that only the Illustrative Plan submitted



8

by the Department of Justice had a lower overall devia­
tion than the plan the court adopted, App., infra, 8a-9a 
& n.4,"‘ but the court rejected that plan because it “split 
numerous counties outside of the metropolitan Atlanta 
area, apparently for racial reasons.” App., infra, 8a n.4. 
The Illustrative Plan in fact split only two counties out­
side the Atlanta area. See App., infra, 44a (map).

The court conceded that “there is some ^population] 
deviation resulting from our adherence to Georgia’s tradi­
tional districting principles and unique factors.” App., 
infra, 8a. The court identified those factors as the State’s 
“strong historical preference for not splitting counties out­
side of the metropolitan Atlanta area,” ibid.; see also App., 
infra, 15a-16a, and the “maintenance of district cores and 
communities of interest,” App., infra, 10a; see also App., 
infra, 17a-18 a. The court acknowledged that “Georgia 
has historically sought to draw district lines so incumbents 
remain in their districts in a new plan and to avoid plac­
ing two incumbents in the same district.” App., infra, 18a; 
see also App., infra, 14a n.lO. The court stated, however, 
that, although it “considered this factor,” it “subordinated 
it to the other considerations” because “it is inherently 
more political than factors such as communities of inter­
est and compactness.” App., infra, 18a. Accordingly, the 
court’s plan places two incumbents—the African-American 
Members of Congress from the former Eleventh and Sec­
ond Districts—in districts with other incumbents, and it 
moves the incumbent from the former Eighth District into 
the Second District. App., infra, 21a.

b. The court also discussed the application of the Vot­
ing Rights Act to its plan, noting that “[djuring the rem­
edy phase of the hearing, a significant portion of the

 ̂This statement was incorrect; in addition to the Illustrative 
Plan, two other plans submitted by the Abrams Interveners had 
lower overall deviations (0.29 and 0.34) than the court’s plan. See 
R. 269 and 319.



argument centered on whether the Court’s remedy would 
be required by Section 2 to contain two maj ority-minority 
districts.” App., infra, 19a. In order to find that a dis­
tricting plan violates Section 2, a court must conclude 
that the minority population is sufficiently compact and 
numerous to form a district, that the minority group is 
politically cohesive, and that whites voting as a bloc 
usually defeat the candidate of choice of minority voters. 
See Thornburg v. Gingles, 478 U.S. 30, 50-51 (1986). 
The latter two factors are referred to collectively as the 
presence of “legally significant racially polarized voting.” 
The court concluded that Section 2 does not require a 
second maj ority-minority district in Georgia “due pri­
marily to the geographic dispersion of [Georgia’s] minority 
population and lack of any significant vote polarization.” 
App., infra, 20a.

With respect to compactness and numerosity, the court 
noted that, “i[i]f Georgia had a concentrated minority pop­
ulation large enough to create a second majority-minority 
district without subverting traditional districting principles, 
the Court would have included one since Georgia’s legis­
lature probably would have done so.” App., infra, 21a 
n.l6. Regarding polarization, the court repeated the 
assertion it had made in its earlier decision in this case 
that, “while some degree of vote polarization exists, it is 
‘not in alarming quantities.’ ” App., infra, 23a (quoting 
864 F. Supp. at 1390). Nonetheless, the court held that 
Section 2 “does require maintenance of the Fifth District 
as a maj ority-minority district,” App., infra, 24a-25a, and 
that “maintaining the percentage of black registered voters 
[in the Fifth District] as close to fifty-five percent as pos­
sible was necessary, in our view, to avoid dilution of the 
Fifth District minorities’ rights.” App., infra, 25a-26a.

c. Judge Edmondson noted his dissent, but did not 
file an opinion. App., infra, 29a.



10

THE QUESTIONS RAISED ARE SUBSTANTIAL

The United States seeks summary reversal of the dis­
trict court’s order imposing upon the State of Georgia a 
congressional redistricting plan that reduces from three 
to one the number of majority-minority districts. The 
plan, which purports to remedy the constitutional defects 
found in the 1992 plan, violates the Constitution, the Vot­
ing Rights Act, and other limitations on court-ordered redis­
tricting. Specifically, the court’s redistricting plan violates 
the one-person, one-vote requirement of Article I, § 2 of 
the Constitution and the mandate of Section 2 of the Vot­
ing Rights Act, 42 U.S.C. 1973, of equal opportunity for 
minority voters to participate in the political process and 
elect representatives of their choice. It also violates the 
rule set forth in Upham v. Seaman, 456 U.S. 37 (1982), 
by making more changes in the existing plan than are 
necessary to correct that plan’s constitutional defects; most 
importantly, the court’s plan disregards the State’s policy 
of having at least two majority-minority districts— a policy 
embodied in all three districting plans enacted by the 
Georgia legislature since the 1990 census.

The district court’s errors are clear. Prompt reversal is 
warranted in order to ensure that the 1996 Georgia con­
gressional elections are not conducted under a malappor- 
tioned districting plan that violates Section 2 of the Voting 
Rights Act. The existence of plans, such as the Illustra­
tive Plan submitted by the United States, that contain two 
majority-minority districts, that comply with state district­
ing policies, and that contain minimum population devia- 
tons shows that, on remand, the district court can quickly 
adopt a plan that satisfies constitutional and statutory 
requirements.

1. The district court clearly erred in adopting a plan 
that deviates from the principle of population equality 
more than is necessary or justifiable. Article I of the Con­
stitution requires that congressional districts achieve 
population equality “as nearly as is practicable.” Wesberry



11

V. Sanders, 376 U.S. 1, 7-8 (1964). This requirement is 
close to an absolute standard. Kirkpatrick v. Preisler, 394 
U.S. 526, 527-528 (1969). When a state legislature 
draws congressional districts, “the command of Art. I, 
§ 2 as regards the National Legislature outweighs the local 
interests that a State may deem relevant in apportioning 
districts for [state legislative] representatives.” Karcher v. 
Daggett, 462 U.S. 725, 732-733 (1983).

Where a state legislature draws congressional districts, 
a minimal deviation from absolute population equality 
may be justified by consistently applied state legislative 
policies, such as “making districts compact, respecting 
municipal boundaries, preserving the cores of prior dis­
tricts, and avoiding contests between incumbent Repre­
sentatives.” Karcher, 462 U.S. at 740. But, as the district 
court correctly noted, App., infra, 7a, the standard for 
court-ordered districting plans is a stricter one. Connor v. 
Finch, 431 U.S. 407, 413-414 (1977); Chapman v. 
Meier, 420 U.S. 1, 26-27 (1975). Because redistricting 
is “ ‘primarily a matter for legislative consideration and 
determination,’ * * * a state legislature is the institution 
that is by far the best situated to identify and then recon­
cile traditional state policies within the constitutionally 
mandated framework of substantial population equality.” 
Connor, 431 U.S. at 414-415 (quoting Reynolds v. Sims, 
377 U.S. 533, 586 (1964)). Courts, on the other hand, 
have no mandate “to compromise sometimes conflicting 
state apportionment policies in the people’s name.” Con­
nor, 431 U.S. at 415.

The district court’s plan contravenes the principle of 
population equality. The largest overall population devia­
tion in the court’s plan is 0.35%, between the Third Dis­
trict (which deviates by +0.12% from absolute population 
equality) and the Second District (which deviates by 
~0.23% ). App., infra, 43a. In contrast, the largest over­
all population deviation in the Illustrative Plan, submitted 
by the Department of Justice, is 0.19%—between the



12

Tenth District (which deviates by +0.09% ) and the 
Eleventh District (which deviates by -0 .10% ). Id. at 
45 a. That difference in overall deviations is sufficient to 
have required the district court to adopt the Illustrative 
Plan (or some other plan with an equal or smaller overall 
deviation), unless there is a compelling justification for 
its failure t0‘ do sO'.

In a footnote, the court justified its decision to adopt a 
plan with a greater population deviation than that in the 
Illustrative Plan on the ground that the Illustrative Plan 
“split numerous counties outside of the metropolitan 
Atlanta area, apparently for racial reasons.” App., infra, 
8a n.4.® That justification is clearly erroneous. The Illus­
trative Plan splits two counties outside the Atlanta area; 
both divisions are explicable on non-racial grounds and 
both are consistent with traditional districting principles.

The Illustrative Plan splits Muscogee County in south­
western Georgia, in which part of Fort Penning is located, 
between the Second and Third Districts. Those districts 
are the two with the highest deviations (-0 .23%  and 
+ 0.12%) in the court’s plan, adding up to an overall de­
viation of 0.35%. By splitting Muscogee County, the 
Illustrative Plan dramatically lowered the deviation from 
population equality of the two districts; the deviation be­
tween the two districts in the Illustrative Plan is, respec­
tively, 0.01 % and 0.02%. In splitting Muscogee County,

® In that footnote, the court also stated that “ [w]hether the 
DOJ should be involved in this, the map-generation phase of the 
redistricting process is questionable,” apparently because “[n]or- 
mally, Georgia would originate a plan for DOJ preclearance under 
[Voting Eights Act] Section 5.” App., infra, 8a n.4. There is no 
reason why the Department of Justice should not be permitted to 
introduce plans in a redistricting case in order to provide bases 
for analysis under Section 2 and constitutional one^person, one-vote 
requirements. Indeed, submission of such a plan is ordinarily essen­
tial in order to make out the preconditions for a Section 2 “results” 
case under Thornburg v. Gingles, 478 U.S. 30 (1986). In any event, 
the district court accepted the submission of the DOJ plan and 
considered it on its merits.



13
the Illustrative Plan also keeps all of Fort Penning— 
\vhich is located in Muscogee and Chattahoochee Coun­
ties—in the Second District. The State’s first post-1990 
plan also kept Fort Penning in one district by splitting 
a county (in that plan, Chattahoochee County). See 
App., infra, 46a (map). The court’s plan keeps Musco­
gee and Chattahoochee Counties intact, but it splits Fort 
Penning between the Second and Third Districts. The 
Illustrative Plan is thus more consistent with the State’s 
own preferences in this area of the State than the Court’s 
plan, results in a very much smaller population deviation, 
and is completely explicable on non-racial grounds (the 
desire to keep Fort Penning in a single district).

The Illustrative Plan also splits Pibb County, in the 
center of the State, among the Third, Eighth, and Elev­
enth Districts. The deviation in those districts under the 
court’s plan is +0.12%, -0 .17% , and +0.08%, respec­
tively. App., infra, 43a. Under the Illustrative Plan, the 
deviation in those three districts is +0.03%, +0.05%, 
and -0 .10% . Id. at 45a. The Illustrative Plan thus 
reduces the deviation very substantially in two of the 
districts and increases it only very slightly in the third.® 
In addition, each of Georgia’s three post-1990 plans also 
split Pibb County. The Illustrative Plan thus follows 
state policies in this respect as well.

® The population deviation of 0.19% between the Tenth and 
Eleventh Districts in the Illustrative Plan could be decreased with 
some minor changes in the Atlanta area or in Bibb County. In the 
course of the proceedings below, the United States endorsed plans 
that contained two majority-minority districts having overall popu­
lation deviations larger than that in the court’s plan. For example, 
we told the court in October that two such plans appeared tO' “com­
ply with the United States Constitution and the Voting Eights Act,” 
Unied States’ Brief on Remedy, R. 348, at 1-2, and we again 
endorsed those plans for their balancing of race and traditional 
districting criteria at the October 30-31 hearing. Tr. 414, 421-422. 
There was testimony, however, that with minor changes, the popula­
tion deviations in those plans could be made much smaller. Tr. 301- 
303 (Selwyn Carter).



14

Keeping county boundaries intact cannot be an abso­
lute goal in the redistricting process. See, e.g., Kirkpatrick 
V. Preisler, 394 U.S. 526, 533-534 (1969) (“[W]e do not 
find legally acceptable the argument that variances are 
justified if they necessarily result from a State’s attempt 
to avoid fragmenting political subdivisions by drawing 
congressional district lines along existing county, munici­
pal, or other political subdivision boundaries.”); Wells v. 
Rockefeller, 394 U.S. 542, 546 (1969) (same); White v. 
Weiser, 412 U.S. 783, 791 (1973) (same)." The sanctity 
of county boundaries is particularly questionable in Geor­
gia. The three-judge court took judicial notice at an 
earlier stage of this case that “until 1962 the county unit 
system was used to undermine the voting strength of 
counties with large black populations.” App., infra, 40a. 
The county unit system was also employed tO' avoid one- 
person, one-vote requirements. See also Wesberry v. 
Sanders, 376 U.S. 1 (1964). The State of Georgia has 
spht counties outside the Atlanta area for purposes of 
congressional redistricting in the past; in the 1970s, 
Whitfield County was divided in a highly irregular pat­
tern between two congressional districts. See 94-631 
J.A. 79. The State does not “consistently appl[y],” 
Karcher v. Daggett, 462 U.S. at 740, a policy of keeping 
counties intact throughout the State; as the three-judge 
court itself found, Georgia’s “preference” against splitting 
counties has not applied at all to the Atlanta area since 
the county unit system was struck down. App., infra, 8a- 
10a. Both the 1972 and 1982 districting plans split coun­
ties in the Atlanta area, and the three-judge court’s own

Both Chapman V. Meier, 420 U.S. 1 (1975), and Connor v. 
Finch, 431 U.S. 407 (1977), dealt with court-ordered state legisla­
tive plans, not congressional redistricting. Accordingly, the district 
court incorrectly relied on Connor, see App., infra, 9a, to justify 
rigid adherence to the asserted state policies against splitting 
counties outside the Atlanta area.



15
plan splits six counties in the immediate area of Atlanta. 
App., infra, 15a.®

2. The district court’s plan violates Section 2 of the 
Voting Rights Act. The Illustrative Plan demonstrates 
that the African-American population in Georgia is suffi­
ciently large and compact—employing the State’s own 
traditional districting criteria—to form a second majority- 
minority district, located in the east central area of the 
State. Voting is racially polarized in that area, as shown 
by evidence in this case and confirmed by the evidence 
upon which the district court based its own finding that 
Section 2 requires drawing a majority-minority district in 
the Atlanta area. The totality of the circumstances— 
including the fact that Georgia’s African-American popu­
lation constitutes 27% of its population and that the 
court’s plan has only one majority-black district out of 
eleven (9 % )—demonstrates that African-Americans will 
not have an equal opportunity to participate in the elec­
toral process and to elect candidates of their choice under 
the court’s plan. We do not contend that a majority- 
minority district must be drawn wherever it is possible 
to do so. The district court’s failure to draw such a 
district in this case, however, violated the substantive 
standards of Section 2. It therefore must be reversed.

a. Compactness/Numerosity. According to the district 
court, a second majority-minority district cannot be drawn 
in Georgia “due primarily to the geographic dispersion 
of its minority population.” App., infra, 20a. That is 
clearly incorrect.®

® Although the court stated that it split those counties “for 
population density reasons,” App., infra, 15a, it later retreated to 
the more modest claim that “it would be impossible to avoid 
splitting any counties,” id. at 18a n.l2.

® The district court made the same finding in its earlier decision 
in this case. See 864 F. Supp. at 1390. That finding was not 
affirmed on appeal because this Court had no occasion to consider 
whether Section 2 required the creation of two majority African- 
American districts in Georgia.



16
The African-American population in east central Geor­

gia is sufficiently large to create a second district. The 
Eleventh District in the Illustrative Plan is located in 
that area. Its population is 54.60% African-American, 
and its VAP is 51.04% African-American. The district 
has a population deviation of only 0.10%. See App., 
infra, 45a.

The Eleventh District in the Illustrative Plan is also 
compact. It is contiguous and its boundaries are regular. 
See J.A. 44a (map). The most appropriate measure of 
compactness of a district is the extent tO' which the district 
comports with the State’s traditional districting practices. 
Jeffers v. Clinton, 730 F. Supp. 196, 207 (E.D. Ark. 
1989), aff’d mem., 498 U.S. 1019 (1991); Houston v. 
Lafayette County, 56 F.3d 606 (5th Cir. 1995). The 
Eleventh District in the Illustrative Plan comports with 
those practices, as identified by the district court.

The district court stated that maintaining counties in­
tact outside the Atlanta area was a traditional practice 
in Georgia. See App., infra, 14a-15a; but see pp. 14-15, 
supra. The Eleventh District in the Illustrative Plan di­
vides only one county outside the Atlanta area, and the 
Illustrative Plan as a whole divides only one additional 
county outside the Atlanta area. See pp'. 12-13, supra

The district court correctly noted that, although its own plan 
splits precincts in some places, App., infra, 9a n.6, Georgia gen­
erally seeks to avoid splitting precincts. App., infra, 9a. The 
Illustrative Plan was built on census tracts, not precincts, and the 
United States did not have data on Georgia’s precincts in its 
mapping program. Precinct lines throughout the State change fre­
quently. Therefore, the precise number of precincts that are split 
in the Illustrative Plan is uncertain. However, since the plan as a 
whole splits only two counties outside the Atlanta area and since 
precinct lines follow county lines, it is extremely doubtful that the 
plan splits an inordinate number of precincts, and the plan could 
certainly be refined not to do so. The plaintiffs-appellees’ unsup­
ported contention that “[t]he DOJ plan * * * splitfs] hundreds of 
precincts,” thereby rendering it “without any value,” is without 
foundation. See Appellees' Brief In Opposition To Application To



17

The district court also identified maintaining the cores of 
existing districts and keeping the State’s four “comer dis­
tricts” intact as traditional state practices. App., infra, 
16a-17a, 17a-18a. The Illustrative Plan recognizes the 
cores of the existing districts and maintains the same 
three corner districts as does the court’s plan. See id. at 
44a (map).“

The Eleventh District in the Illustrative Plan does not 
suffer from the defects in the district that this Court found 
unconstitutional in Miller v. Johnson. The Illustrative 
Plan’s Eleventh District does not incorporate the Athens- 
Augusta strip that the district court found was “histori­
cally” located in the Tenth District, see App., infra, 14a, 
nor does it contain any part of the southeastern coastal 
district, as did the district that this Court found uncon­
stitutional in Miller, see id. at 19a. As the Illustrative 
Plan therefore shows, the African-American population 
in DeKalb County and central Georgia is “sufficiently 
large and geographically compact to constitute a major­
ity” in a congressional district, Thornburg v. Gingles, 
478 U.S. at 50, without subordinating “Georgia’s tradi­
tional districting policies,” “constitutional norms,” or 
“common sense,” App., infra, 21a.

Justice Kennedy, Circuit Judge, For A Stay and./or Injunction 
Pending Appeal 8 & n.7.

The district court observed that its plan “maintains ninety- 
five counties (totally or partially) in the same districts as they 
were located in the 1982 plan—significant given the addition of 
the Eleventh District.” App., infra, 10a. However, under the same 
measure, the Illustrative Plan fares better; it maintains 101 
counties totally or partially within the same districts as in the 
1982 Plan. Similarly, when compared to the plan the Georgia 
Assembly first passed in 1991— p̂rior to any objection from the 
United States Attorney General— t̂he Illustrative Plan maintains 
138 of Georgia’s 159 counties totally or partially in the same 
districts while the district court’s plan keeps only 123 counties in 
the same districts. Compare 94-631 J.A. 80 with App., infra, 44a 
(map).



18
b. Racially Polarized Voting. The evidence in the 

record in this case established the existence of legally 
significant racially polarized voting. The district court, 
however, at one point referred to what it termed the “lack 
of any significant vote polarization.” App., infra, 20a. 
The court also stated that “some degree of [racial] vote 
polarization [in Georgia] exists” but “not in alarming 
quantities,” and that “[t]he second and third Gingles pre­
conditions therefore would not be met in a Section 2 chal­
lenge to the [court’s] remedy.” App., infra, 23a (quoting 
864 F. Supp. at 1390). To the extent those statements 
are read as factual findings that racially polarized voting 
in Georgia—and, in particular, in the east central portion 
of the State—does not satisfy the second and third 
Gingles preconditions, they are clearly erroneous.

The district court appeared to base its statements re­
garding racially polarized voting on the findings it made 
at the earlier stage of this case. At that time, the court 
found that 22% to 38%> of white people in the area of 
the 1992 Eleventh District (which substantially overlaps 
the area of the Eleventh District in the Illustrative Plan) 
would vote for African-American candidates, while 20% 
to 23% of African-Americans in the area would vote for 
white candidates. 864 F. Supp. at 1390.

Those findings (which, incidentally, substantially over­
state the degree of white crossover voting ®̂) are sufficient

12 Experts for the United States and plaintiffs found that white 
crossover voting ranged from 0% to 27%, and that black cross­
over voting ranged from 3% to 11% in the most relevant elections.
U. S. Exh. 24, tables 1 & 3; PX 82, Attachment E. The court’s 
higher crossover figures included consideration of judicial elections 
involving an appointed black incumbent and elections in which there 
were multiple black candidates and a white candidate. Both kinds 
of elections are substantially less probative of racially polarized vot­
ing than ordinary elections between one black and one white candi­
date. See 5 T. Tr. 203-204, 228 (testimony of Allan Lichtman). 
See also Thornburg V. Gingles, 478 U.S. 30, 57 (1986) ; Collins
V. City of Norfolk, 883 F.2d 1232, 1243 (4th Cir. 1989), cert, 
denied, 498 U.S. 938 (1990). The plaintiffs’ expert. Dr. Weber,



19

to demonstrate the existence of legally significant racially 
polarized voting in the district court’s plan, especially in 
the area in which the Illustrative Plan’s Eleventh District 
is located. The court’s plan contains three districts—the 
Fourth, the Tenth, and the Eleventh—that each contain 
parts of the area covered by the Eleventh District in the 
Illustrative Plan. Of those three districts, the Tenth Dis­
trict has the largest African-American VAP—approxi­
mately 34%. Calculating white crossover voting at 30% 
and black crossover voting at 22%—the midpoints of the 
ranges found by the district court—a black candidate 
would be unable to win election in the court’s Tenth 
District.^® The same result would follow even more 
clearly in the court’s Fourth and Eleventh districts, which

testified that judicial elections do not provide a good indication of 
voting patterns in congressional elections (4 T. Tr. 318) and that, 
if the judicial incumbent elections were removed from his own 
calculation, the white crossover vote would drop to 26.&%. 4 T. Tr. 
324-325. If the elections involving multiple black candidates were 
removed, he testified that the white crossover calculation would 
drop to 17,%. 4 T. Tr. 325.

The figure for black political cohesion is the inverse of the 
figure for black crossover voting. Under the assumption stated in 
the text of 22% black crossover voting, the figure for black politi­
cal cohesion is 78% (100%-22%). The vote for the minority- 
preferred candidate can be derived by multiplying the black VAP 
(34%) by the figure for black political cohesion (78%), and then 
adding the white VAP (66%) multiplied by the percentage of 
white crossover voting (30%). Under the assumptions in the 
text, and assuming also that turnout rates for the two races are 
identical, the black-preferred candidate would receive 46% of the 
vote (and the white-preferred candidate would receive 54%) in 
the court’s Tenth District. If black crossover voting remained 
at approximately 22%, the white crossover vote would have to 
be approximately 36%—or at the very top of the range found by 
the district court—in order for a black-preferred candidate to win. 
If the consistenly lower black turnout rates found by the district 
court are factored in, see 864 F. Supp. at 1392, the white crossover 
vote would have to be still higher for a minority-preferred candidate 
to win.



20

have African-American VAPs of 33% and 11%, respec­
tively. Voting polarization will thus prevent African- 
American voters in this area of the State from electing 
candidates of choice under the district court’s plan. The 
Illustrative Plan’s Eleventh District would remedy this 
discrimination.

At trial, the parties stipulated to the facts that, with one 
exception, all of the African-American state legislators in 
Georgia were elected in districts that were majority Afri­
can-American and that all but one of the 181 majority- 
white state legislative districts were represented by white 
legislators. See 94-631 J.A. 26-27 (fll 61-64). Those 
facts support the conclusion that Georgia suffers from 
racially polarized voting, resulting in the usual defeat 
of minority-preferred candidates in majority-white election 
districts.

The expert testimony reached the same conclusion. The 
United States’ expert, Dr. Lichtman, found that white 
crossover voting ranged from 0% to 27 %. See note 12, 

supra. The plaintiffs’ expert, Dr. Weber, corroborated the 
results found by Dr. Lichtman with respect to the levels 
of polarization in the area of the Eleventh District. PX 
82, at 22-24 & Attachment E. In each instance, however, 
he found the results lacking in significance because, de­
spite the polarization, the African-American candidates 
won within the borders of the 60% majority VAP African- 
American 1992 Eleventh District. Id. at 24-27. Far from 
casting doubt on the degree of racially polarized voting, 
this testimony supported the showing that must be made 
to establish a Section 2 claim: that candidates of choice 
of African-Americans in the area of the 1992 Eleventh 
District can win majorities only if African-Americans form 
a majority of the constituency.

The district court’s apparent findings of lack of signifi­
cant racial polarization also directly conflict with its 
Section 2 findings regarding the Atlanta area. The court 
found that “Section 2 * * * does require maintenance 
of the Fifth [Atlanta] District as a majority-minority dis-



21

trict.” App., infra, 24a-25a. The court found that polari­
zation there was sufficiently serious that “maintaining the 
percentage of black registered voters as close to fifty-five 
percent as possible was necessary, in our view, to avoid 
dilution of the Fifth District minorities’ rights.” Id. at 
25a-26a. Those findings cannot be squared with the court’s 
statements regarding a supposed lack of racial polarization 
in the Eleventh District.

In making its findings regarding the Fifth District, the 
court relied exclusively on the analysis prepared by Dr. 
Katz, the State’s expert, at the original trial. App., infra, 
25a; see also 864 F. Supp. at 1390, 1391-1392. Dr. 
Katz analyzed a large group of Georgia election results 
and determined the racial composition of the registered 
electorate in each case. From those data, he calculated 
the probability of a black candidate winning an election 
in Georgia as a function of the percentage of blacks in 
the relevant electorate. Katz Report, DX 170, at 18-21; 
5 T. Tr. 49-50, 62-71. He concluded that the probability 
of a black candidate winning an election in Georgia does 
not go above 50% until the percentage of registered 
voters who are black reaches approximately 50%. 5 T. 
Tr. 84-85.

Dr. Katz’s study supports— and thus tends to corrobo­
rate—the conclusion that African-Americans can win 
elections in Georgia only when they are in the majority 
in the constituency.^^ Moreover, the study is based on 
statewide data. The study contained no evidence that 
polarization might differ in the areas of the Fifth and 
Eleventh Districts. If the study demonstrates that there 
was racially polarized voting in the area of the Fifth Dis-

«  Dr. Katz’s study does not directly address the question whether 
legally significant racially polarized voting exists, since it does not 
make “discrete inquiries into minority and white voting practices.” 
Gingles, 478 U.S. at 56. In addition, Dr. Katz’s “statistics are based 
on statewide figures.” App., infra, 26a n.l8. It is doubtful that 
exclusive reliance on such statewide figures would be appropriate in 
a district-specific Section 2 challenge. 478 U.S. at 59 n.28.



22

trict, it therefore also demonstrates that there is similar 
racially polarized voting in the east-central area of the 
State, where the Illustrative Plan’s Eleventh District is 
located. The court’s contrary conclusions with regard to 
the two areas are therefore internally inconsistent.

c. All the other factors that contribute to a “totality 
of circumstances” analysis establish that the district court’s 
failure to draw a second majority-black district in central 
Georgia violates Section 2. The district court’s finding 
that Section 2 requires creation of a majority-black dis­
trict in the Atlanta area confirms that conclusion.

The history of racial discrimination in voting in Geor­
gia has been spelled out in prior decisions, and the dis­
trict court took judicial notice of that history in the initial 
proceedings in this litigation. See App., infra, 39a-40a. 
From the time of Georgia’s first constitution, which barred 
blacks from voting as a matter of law, Georgia has em­
ployed a variety of devices to destroy or dilute black 
voting strength. The State has employed literacy tests, 
white primaries, and the county-unit system. The State 
has also used districting as a way of discriminating 
against African-American citizens. Ibid. From Recon­
struction until the 1980s, black voters were not in the 
majority in any of the State’s congressional districts. In 
1982, the District Court for the District of Columbia 
found that the State of Georgia had acted with a dis­
criminatory purpose in adopting a congressional redistrict­
ing plan that failed to include a district in the Atlanta 
area whose registered voters were majority African- 
American. See Busbee v. Smith, 549 F. Supp. 494 
(D.D.C. 1982), aff’d, 459 U.S. 1166 (1983).

At the original trial, the parties stipulated to Georgia’s 
history of past discrimination. See Joint Statement of 
Undisputed Facts, 76-98, 99-101, 104-134. Among the 
consequences of that history are continuing and signifi­
cant socioeconomic disparities between black persons in 
Georgia, as a group, and white persons as a group. See 
94-631 J.A. 29-33 ( f f  135-155). Those differences re-



23
suit in African-Americans having a smaller economic base 
on which to draw to mount political campaigns. In addi­
tion, the evidence showed that African-Americans have 
lower registration and turnout rates than whites. See 
U.S. Exh. 24, at 19, 21-23. The political handicaps suf­
fered by blacks in majority-white political districts are 
aggravated by the large geographical area covered by 
Georgia’s congressional districts and by Georgia’s require­
ments that those who do not win a majority vote in a 
primary face runoff elections. See Thornburg v. Gingles, 
478 U.S. 30, 45 (1986). Finally, the district court’s 
plan, with only one majority-black district, would leave 
black voters substantially underrepresented in the State’s 
congressional districting plan, while creating two majority- 
minority districts out of eleven (18% ) would not over­
represent blacks, who constitute 27% of the State’s popu­
lation. See Johnson v. DeGrandy, 114 S. Ct. 2647, 2662- 
2663 (1994).

3. The discussion in Part 2(a) above demonstrates as 
well that the district court departed unduly from the 
State’s own districting policies, in contravention of Upham 
v. Seamon, 456 U.S. 37 (1982). In that case, the Court 
observed that “i[w]henever a district court is faced with 
entering fa] reapportionment order, * * * it is faced with 
the problem of reconciling the requirements of the Con­
stitution with the goals of state political policy.” Id. at 
43 (internal quotation marks omitted). That reconcilia­
tion “can only be reached if the district court’s modifica­
tions of a state plan are limited to those necessary to cure 
any constitutional or statutory defect.” Ibid. As the dis­
trict court stated in this case, “[i]n fashioning a remedy 
in redistricting cases, courts are generally limited to cor­
recting only those unconstitutional aspects of a state’s 
plan.” App., infra, 2a.

The district court recognized that state legislative policy 
dictated the creation of two majority-minority districts, in­
sofar as that could be accomplished within constitutional 
and statutory limits and without unduly departing from



24

other traditional districting principles. The court stated 
that “l[i]f Georgia had a concentrated minority popu­
lation large enough to create a second majority-minority 
district without subverting traditional districting principles, 
the Court would have included one since Georgia’s legis­
lature probably would have done so.” App., infra, 21a 
n.l6.

The Illustrative Plan shows that, using that test, the 
district court should have adopted a plan with two 
majority-minority districts. It demonstrates that a plan 
could have been drawn that comports with Georgia’s tra­
ditional geographic districting practices and that creates 
two compact majority-minority districts.“  Indeed, the

1̂5 The court correctly assessed the state legislative policy in this 
area. The first plan that the State adopted in 1991, prior to either 
of the Section 5 objections that were lodged by the Department 
of Justice, provides a useful guide to the State’s districting princi­
ples. Although the district court stated that the 1991 Plan “con­
tained many of the same features in its Eleventh District that 
found their way into the unconstitutional Eleventh,” App., infra, 
13a n.9, the court did not find that the 1991 Plan was itself 
unconstitutional or the product of any unconstitutional motive. The 
1991 Plan included two majority-minority districts, one of which 
(the Eleventh) was located in the same general area of the State 
and based on the same DeKalb County tO' Bibb County axis as the 
Eleventh District in the Illustrative Plan. Compare App., infra, 
46a (1991 Plan) with App., infra, 44a (Illustrative Plan). The 
State eventually drew the district in a way that led to its being 
found unconstitutional. Id. at 13a n.9. That does not negate the 
fact that, from the start, it was part of the State’s redistricting 
policy to draw two majority-minority congressional districts in 
Georgia, including one in the area of the Eleventh District.

1® The district court’s rejection of the State’s decision to draw 
a majority-black, urban/suburban Eleventh Congressional District 
stretching southeast from metropolitan Atlanta in favor of a 
majority-white, urban/suburban district stretching east from metro­
politan Atlanta is particularly troubling, because it treats the 
State’s efforts to create a majority-black district “less favorably 
than similar efforts on behalf of other groups.” See Miller v. 
Johnson, 115 S. Ct. at 2497 (O’Connor, J., concurring).



25

Illustrative Plan would better comport with the State’s 
“historid[]” policy of protecting incumbents, App., infra, 
18a; see also id. at 14a, than does the Court’s plan, 
which pits incumbents against each other in two districts 
and moves an additional incumbent to a different district, 
see p. 8, supra. Unlike the Court’s plan; the Illustrative 
Plan leaves all incumbents in the districts they currently 
represent. See United States’ Reponse to Plaintiffs’ Plans 
Remedy 4X and 4X-R, R. 369 at 10 (filed Nov. 21, 
1995). Nor does the Illustrative Plan suffer from the 
defects that this Court found in Miller v. Johnson', it does 
not take race into account as the “predominant, over­
riding factor,” Miller, 115 S. Ct. at 2490, that “subordi­
nated traditional race-neutral districting principles,” id. 
at 2488, or substantially disregard “customary and tradi­
tional districting practices,” id. at 2497 (O’Connor, J., 
concurring). There was therefore no basis for the district 
court to disregard the state legislature’s choices.

The district court’s error is substantially similar to the 
error committed by the district court in White v. Weiser, 
412 U.S. 783 (1973). In that case, two remedial plans 
were presented to the district court. The court selected 
the one that had a higher overall population deviation and 
departed more from the malapportioned plan that the 
State had adopted, but that split fewer counties. See 412 
U.S. at 795-796, 797 n.l6. This Court reversed that deci­
sion, finding that the court should have adopted the plan 
with a lower population deviation that “most clearly ap­
proximated the reapportionment plan of the state legisla­
ture, while satisfying constitutional requirements.” Id. at 
796. In Upham v. Seamon, 456 U.S. 37, 41-42 (1982), 
the Court relied on that result to reverse summarily a 
district court’s adoption of a remedial plan that disregarded 
State policy more than necessary to cure the constitu­
tional violation. The same disposition is warranted here.



26

4. As described above, the district court’s plan vio­
lates the constitutional one-person, one-vote requirement. 
Section 2 of the Voting Rights Act, and the requirement 
that a court-ordered districting plan depart as little as 
possible from the State’s legitimate districting choices. 
Although Georgia has not, since the 1990 census, enacted 
a congressional districting plan with only one majority- 
minority district, the district court has adopted such a 
plan, despite the demonstration to it that a plan with two 
compact majority-minority districts could have been 
adopted, and that such a plan could have smaller popu­
lation deviations than the court’s plan.

If the court’s plan is allowed to go into' effect, it will 
cause irreparable injury to residents of overpopulated 
districts in general, and to African-American voters in 
particular; both groups will be underrepresented for the 
next two years. In addition, a severe burden would be 
imposed on Georgia’s electorate as a whole if the 1996 
elections are held under a plan that is likely to be found 
invalid and consequently replaced with another new 
plan—the third since 1990—in 1998.

As we have shown above, the outlines of a legally valid 
plan containing two compact majority-minority districts 
and minimum population deviations are reasonably clear. 
The case should be remanded to the district court with 
instructions to adopt such a plan. Although candidate 
qualification would have to be rescheduled,”  we are 
confident that, if the Court acts promptly and the district 
court acts thereafter “with a compelling awareness of the 
need for * * * expeditious accomplishment,” Connor v. 
Finch, 431 U.S. 407, 426 (1977), an acceptable plan can 
be put into place with minimal disruption to this year’s 
election schedule.

Candidate qualification is scheduled for April 22-26, 1996, the 
first-round primary elections are scheduled for July 16, and run-off 
elections are scheduled for August 6.



27

CONCLUSION
This Court should summarily reverse the district court’s 

order and remand the case to the district court with in­
structions to develop and put in place a new plan that is 
consistent with constitutional one-person, one-vote require­
ments and that contains two compact majority-minority 
districts. In the alternative, this Court should note prob­
able jurisdiction and set the case for full briefing and 
argument.

Respectfully submitted.

Drew S. D ays, III
Solicitor General

D eval L. P atrick 
Assistant Attorney General

P aul Bender 
Deputy Solicitor General

James A. F eldman 
Assistant to the Solicitor General

Steven H. Rosenbaum  
Miriam R. Eisen stein  

Attorneys
March 1996



APPENDIX A

[Filed Dec. 18, 1995]

UNITED STATES DISTRICT COURT 
SOUTHERN DISTRICT OF GEORGIA 

AUGUSTA DIVISION

CV 194-008

Davida Jo h n so n ; Pam  B u r k e ; 
H enry  Zittrouer; George L. D eloach ; 

and George Seaton, plaintiffs

V.

Zell Miller, in his official capacity as Governor 
of the State of Georgia; et al ., defendants

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

ORDER
This Court’s September 12, 1994, Order declared 

Georgia’s Eleventh Congressional District unconstitu­
tional. Johnson v. Miller, 864 F. Supp. 1354 (S.D. 
Ga. 1994) {“Johnson I”). The Supreme Court af­
firmed our decision in Miller v. Johnson, ------ U.S.
------, 132 L.Ed.2d 762 (1995) {“Johnson II” ). In
accordance with the Supreme Court’s affirmance, we 
held a hearing on August 22, 1995, to determine the

(la)



2a

best means of resolving the issue of remedy. Plain­
tiffs moved for leave to add residents of Georgia’s 
Second Congressional District in a constitutional chal­
lenge thereto. The Court granted that motion, held 
a trial, and declared the Second District unconstitu­
tional. Johnson v. Miller, ------F. Supp. -------  (S.D.
Ga. 1995) {‘̂ Johnson III” )."

As to the remedy, we deferred to Georgia’s legisla­
ture, allowing it an opportunity to draw a new con­
gressional map in accordance with the Supreme 
Court’s holding and this Court’s prior findings. Growe
V, Emison,------U .S .------- , 122 L.Ed.2d 388 (1993).
Time obviously was of the essence in light of the 
rapidly approaching 1996 congressional elections and 
the more imminent 1996 congressional campaigns. 
However, the legislature notified the Court on Sep­
tember 13, 1995, that it was unable to redraw the 
map and had adjourned, effectively leaving the task 
to us. See “Notice of Def. Murphy with Respect to 
Legislative Adjournment.” We are therefore forced 
to redraw Georgia’s congressional districting plan.

I. Scope of Remedy
In fashioning a remedy in redistricting cases, 

courts are generally limited to correcting only those 
unconstitutional aspects of a state’s plan. Upham v. 
Seamon, 456 U.S. 37 (1982). The rationale for such 
a “minimum change” remedy is the recognition that 
redistricting is an inherently political task for which

* All page references to Johnson III are to the manuscript 
Order on file with the Clerk’s office. All references to tran­
scripts from the Eleventh District trial will be “Eleventh 
Trl. Trans.,” and reference to the Second District trial will be 
“Second Trl. Trans.”



3a

federal courts are ill-suited. Id. at 41-42, A mini­
mum change plan acts as a surrogate for the intent of 
the state’s legislative body. See Id. at 43 (district 
court must reconcile Constitutional requirements with 
goals of state policy); White v. Weiser, 412 U.S. 783, 
795 (1973) (“district court should . . . honor state 
policies in the context of congressional reapportion­
ment”).

In Upham, the Texas legislature had passed a con­
gressional districting plan and submitted it to the 
U.S, Department of Justice (“DOJ” ) for preclear­
ance. While the plan was with the DOJ, a case v/as 
filed in district court challenging the plan’s constitu­
tionality under the Voting Rights Act (“VRA” ). The 
U.S, Attorney General also objected to the plan, spe­
cifically objecting to two of the plan’s twenty-seven 
districts. The three judge panel determined that the 
two districts were unconstitutional, and devised a 
plan of its own. The court’s plan redrew the two 
objectionable districts and two adjoining districts. 
However, in doing so, the Court substituted its own 
judgment for the Texas legislature’s and redrew dis­
tricts in Dallas County in addition to the two objec­
tionable south Texas districts. Reasoning that the 
legislatively-drawn districts would have failed retro­
gression analysis, the court redrew the Dallas County 
districts under the stricter standard applicable to 
courts. Thus, the court ignored the legislature’s pref­
erences.

Reversing, the Supreme Court held that the dis­
trict court erred in redrawing the Dallas County dis­
tricts absent an objection to those by the attorney 
general or a specific finding that they were uncon­
stitutional. The Court stated that a “district court’s 
modifications of a state plan [should be] limited to



4a

those necessary to cure any constitutional or statu­
tory defect.” Upham, 456 U.S. at 43. And, “ ‘[i]n 
fashioning a reapportionment plan or in choosing 
among plans, a district court should not pre-empt the 
legislative task nor intrude upon state policy any 
more than necessary,’ ” Id. at 42 (quoting Weiser, 
412 U.S. at 794-95). The Court had no objection to 
the district court’s correction of the two objectionable 
districts or the two adjoining districts that were nec­
essarily impacted as a result thereof.

At first glance it appears that, in remedying Geor­
gia’s constitutionally infirm congressional districting 
plan, we are faced with the same task as the Upham 
court, which would entail making only those changes 
necessary to bring the current plan into constitutional 
compliance. However, this Court’s remedial task dif­
fers from that of the Upham court’s because Geor­
gia’s current plan was not the product of Georgia’s 
legislative will. Eather, the process producing Geor­
gia’s current plan was tainted by unconstitutional 
DOJ interference. See Johmon I, 864 F. Supp, at 
1367-68. By virtue of its unconstitutional origin, 
Georgia’s current congressional plan cannot form the 
basis for the remedy we now construct because it does 
not represent the goals of Georgia’s historic policies 
nor the state legislature’s true intent.

A. Department of Justice’s Interference
Central to this Court’s finding that the Second and 

Eleventh Districts are unconstitutional was the DOJ’s 
level of involvement in the creation of the current 
plan. The DOJ used informants inside the legislature 
to keep tabs on the legislature’s progress. Id. at 1367. 
It worked closely with the ACLU to help the ACLU



5a

achieve its objective of three majority-minority dis­
tricts in Georgia. Id. at 1362-63 (finding, among 
other things, that the “DOJ "was more accessible and 
amenable to the opinions of the ACLU than to those 
of the Attorney General of the State of Georgia” ). 
The DOJ rejected Georgia’s first two plans even 
though they clearly did not violate Section 5 of the 
Voting Rights Act. Johnson II, 132 L.Ed.2d at 784. 
The DOJ basically used the preclearance process to 
force Georgia to adopt the ACLU redistricting plan 
and, in the process, subvert its own legislative pref­
erences to those of the DOJ. The product of the proc­
ess, the unconstitutional plan, more closely reflects 
the DOJ’s and ACLU’s intentions than the Georgia 
General Assembly’s. Johnson I, 864 F. Supp. at 1367 
(finding the precleared plan “bore all the signs of 
DOJ’s involvement”) . Using the current plan as a 
basis for the remedy would, in effect, validate the 
DOJ’s constitutionally objectionable actions. Thus, 
we cannot use Georgia’s current plan as a surrogate 
for the legislature’s reapportionment policies and 
goals. See Johnson II, 132 L.Ed.2d at 782 (Georgia’s 
. . . redistricting plan cannot be upheld unless it satis­
fies strict scrutiny,” which the Court found it did 
not) (emphasis added); Johnson I, 864 F. Supp. at 
1393 (finding “ [i]n sum, the current districting plan 
is not reasonably necessary to comply with Sections 
2 or 5 of the VRA. Since no compelling state interest 
other than VRA compliance is evident, the plan fails 
strict scrutiny under the Fourteenth Amendment” ) 
(emphasis added).

Because the current plan does not represent the 
Georgia legislature’s intent, we are not bound by Up- 
ham to make only minimal changes to the current



6a

plan in fashioning a remedy.® Finally, since two of 
Georgia’s districts are unconstitutional, are on op­
posite sides of the state, and contain all or parts of 
nearly a third of Georgia’s counties, any remedy even 
minimally disruptive to the current plan would nec­
essarily have resulted in drastic changes.̂ '* Consider­
ing the objective that each congressional district 
should ideally contain 588,929 persons, it is a virtual 
impossibility to change less than we have.

B. Remedy
Because we are unable to use Georgia’s current 

plan as the basis for a remedy, we were compelled 
to devise our own plan. The remedy we devised and 
now order into effect is attached to this order as Ap­
pendix A, with the verbal census block description at­
tached as Appendix B. The plan contains one 
majority-minority district, has an overall population 
deviation of .35%, and an average deviation of .11%. 
The plan and how we devised it we further explain 
below.

II. Remedial Factors
Since the Court is not limited to Georgia’s cur­

rent unconstitutional plan, the Court’s task is akin 
to those cases in which states had no plans. Thus, 
when devising the remedy, the Court was bound by

 ̂This does not mean that we ignored the current plan 
altogether, only that our consideration of it was tempered by 
its origins.

® The DOJ’s demand for three majority-minority districts 
resulted in land bridges, splitting numerous counties outside 
of metropolitan Atlanta predominantly for racial reasons, and 
splitting 70 voting precincts for the same reasons. Johnson 
I, at 1367; Johnson III, at 6.



7a

the stricter guidelines applicable to court plans. These 
guidelines include the one person-one vote require­
ment and the state’s traditional districting principles.

A. One Person, One Vote
The Equal Protection clause of the Constitution re­

quires “no less than substantially equal . . . repre­
sentation for all citizens, of all places as well as of 
all races.” Reynolds v. Sims, 377 U.S. 533, 568 
(1964). In adhering to this principle of one person- 
one vote, the Supreme Court has allowed legislatures 
only the narrowest of margins in population devia­
tions when drawing congressional districts. See Wes- 
berry v. Sanders, 376 U.S. 1, 7-8 (1964) (requiring 
congressional districts to be “as nearly as is prac­
ticable” to one person-one vote); Karcher v. Daggett, 
462 U.S. 725, 732-33 (1983) (“absolute population 
equality [is] the paramount objective” in congres­
sional apportionment). Since federal courts are held 
to stricter standards than legislatures in redistrict­
ing, Wise V. Lipscomb, 437 U.S. 535, 540 (1978), we 
were particularly constrained to create a remedy with 
the lowest population deviation practicable.

Our remedy’s population deviation is not perfect. 
The Supreme Court’s mandate that courts follow a 
state’s historical legislative districting principles some­
times conflicts with the one person-one vote require­
ment. Where such tensions existed, we sought to de­
fer to one person-one vote without decimating the his­
torical character of Georgia’s congressional plan. 
The resultant remedy contains a lower deviation than 
the current Georgia plan, the 1982 plan, and any 
other plan presented to the Court which was not



8a

otherwise constitutionally defective/ Nevertheless, 
there is some deviation resulting from our adherence 
to Georgia’s traditional districting principles and 
unique factors, some of which we pause to reconcile 
here, the remainder of which we discuss in Part II 
(B) infra.

Georgia’s strong historical preference for not split­
ting counties outside of the metropolitan Atlanta 
area, as well as other unique factors, account for 
some of our plan’s deviation. See Connor v. Finch, 
431 U.S. 407, 419-20 (1977) (in court plan reappor­
tioning state legislature, court is required to enun­
ciate historically significant state policy or unique 
features that result in deviation from ideal district 
size) ; Weiser, 412 U.S. at 795 (1973) (“District 
Court should . . . honor state policies in the context 
of congressional reapportionment”).

Georgia has preferred to split counties only in the 
Atlanta area where division was necessary because of 
Atlanta’s dense population relative to other areas of 
the state. We did rearrange precincts in the Atlanta 
area for purposes of achieving better deviations.® Of

* The DOJ presented a plan containing a slightly lower over­
all deviation than the Court’s plan. However, its plan split 
numerous counties outside of the metropolitan Atlanta area, 
apparently for racial reasons. Whether the DOJ should be 
involved in this, the map-generation phase of the redistricting 
process is questionable. Normally, Georgia would originate 
a plan for DOJ preclearance under VRA Section 5. Regret­
tably, we are standing in the state’s shoes, and our plan is not 
subject to Section 5 in any event.

“ The Court was able to achieve such low deviations by 
placing Talbot County in the Third District, thereby reducing 
the Second District’s deviation to —.23%. This also allowed 
a transfer of population from the Third District to the under-



9a

course, we attempted to avoid splitting voting pre­
cincts (with the exception of Cobb and Clayton 
Counties)* due to Georgia’s preference against that 
practice.’’ The unconstitutional plan splits twenty-

populated Fifth District. The Fifth District does extend into 
Dekalb County somewhat, an extension necessary to allow 
other precincts in the Atlanta area to be relocated in order 
to achieve vastly improved deviation, and also keep the City 
of Atlanta wholly within the Fifth District. The Court’s 
splitting of Dekalb and other Atlanta area counties is con­
sistent with the Georgia General Assembly’s decision to split 
those counties due to the heavy population concentrations in 
that region of the state.

® A look at a map of the City of Marietta and Cobb County 
reveals that every plan drawn at any level (county commis­
sion, state house, state senate, and congressional) must of 
necessity split some precincts in the Marietta-Cobb County 
area because of an annexation pattern that created noncon­
tiguous enclaves and exclaves. There are pieces of the city 
that are islands in the county and vice versa. Def. Exh. 92.

In Clayton County, two precincts were split along already 
existing school board or county commission lines, and a third 
was split along a major geographical feature. These enabled 
the Court to achieve lower deviations in the Third and Fifth 
Districts without causing the problems discussed in footnote 
seven, infra.

As we have noted:
[0,]ne bad side effect of the splitting of units as small as 
precincts for racial purposes occurs when the precinct is 
divided in one way for a state house seat, another way for 
a state senate seat, and yet a different way for a congres­
sional seat. When this kind of gerrymandering is re­
peated often enough, the voting combinations are so 
rare . . . that a secret ballot becomes difficult for some 
voters.

Johnson III at 6 n.3; Second TrI. Trans, at 28. The Court 
believes the same result could obtain from splitting precincts



10a
eight precincts in the Second District alone (seventy 
overall), all predominantly for racial reasons. JoKti- 
son III, at 6. The product of the Court’s efforts is a 
lower overall deviation than Georgia’s 1982 and 1992 
plans.

However, maintaining political subdivisions (coun­
ties) cannot alone justify less than perfect devia­
tion in a court plan. Cf. Karcher, 462 U.S. at 740 
(“we are willing to defer to state legislative policies, 
so long as they are consistent with constitutional 
norms, even if they require small differences in the 
population of congressional districts”) ; Weiser, 412 
U.S. 783 (1973) (mandating strict adherence to leg­
islative policies). We therefore considered splitting 
counties outside of the Atlanta area, but other factors 
unique to Georgia militated against it.

Other factors causing the slight deviation include 
maintenance of core districts and communities of in­
terest. Georgia has historically preferred to maintain 
the core of each district during reapportionment. Cf. 
1972 Plan, Def. Exh. 14, and 1982 Plan, PI. Exh. 
41, Def. Exh. 15. Our remedy maintains ninety- 
five counties (totally or partially) in the same dis­
tricts as they were located in the 1982 plan—signifi­
cant given the addition of the Eleventh District, 
which in the remedy contains thirteen counties.

Maintaining county boundaries also helps to ensure 
that communities of interest are maintained within 
the same district. Residents of a particular county 
are similarly affected by the decisions of their elected 
representatives. People who share communities of

for any reason, not only racial. Thus, except in one instance, 
the Court avoided splitting precincts unless they were split 
in the same manner at other election levels, and then only 
where absolutely necessary. The new split precinct has its 
boundaries along easily identified major roadways.



11a

interest logically belong within the same congres­
sional district. See Johnson II, 132 L.Ed.2d at 782 
(discussing significance of communities of interest).

Incidentally, a factor unique to Georgia, and one 
which enables a draftsman to reach near perfect 
deviations without splitting counties, is the number 
of Georgia counties: 159. Most of the counties are 
more sparsely populated than the Atlanta area coun­
ties. With such a high number of counties, we were 
able to achieve extremely low deviations without split­
ting non-Atlanta area counties. Cf. Mahan v. Howell, 
410 U.S. 315, 321 (1973) (allowing states to “use 
political subdivision lines to a greater extent in es­
tablishing state legislative districts” because of the 
greater number of state legislative districts to be 
apportioned as compared to the lesser number of con­
gressional districts), modified 411 U.S. 922 (1973).'® 
As we have noted:

With the exception of Texas, Georgia has more 
counties than any other state in the Union; one 
would think that such a proliferation would pro­
vide ample building blocks for acceptable voting 
districts without chopping any of those blocks in 
half.

Johnson I, 864 F. Supp. at 1377.

In the calculus of district population deviation, our 
only measure of the state’s demographics is the de­
cennial census. Since the population is not static, we

® The correspondingly high number of counties in Georgia 
afforded the Court the same opportunity to maintain county 
boundaries (outside of metropolitan Atlanta counties) among 
the eleven congressional districts.



12a

adhered to the fiction that the census block figures are 
accurate to the exclusion of all others. In 1995, no 
precise count of a district’s population can be made 
using 1990 data. We are halfway through the decade. 
Changes have occurred in Georgia. See, e.g., footnote 
17, infra. In sum, using the available data and whole 
counties outside of the Atlanta area, our remedy is 
probably as close to ‘‘zero deviation” as one can be. 
At this point in the decade, the deviation effort is 
largely theoretical. The real data, known only to 
providence, would doubtlessly lead us to another re­
sult.

Yet, we remain sensitive to the extant deviation, 
so we now explain the other factors we were com­
pelled to consider.

B. Adherence to Georgia’s Historic Districting Pref­
erences

1. Source of Georgia Legislature’s Intentions
Since redistricting is an inherently political proc­

ess, federal courts undertaking the task must adhere 
to the state’s historical districting preferences and 
traditions. Upham, 456 U.S. at 39; Weiser, 412 U.S. 
at 795. The problem facing the Court was that, un­
like the situation in Upham, Georgia’s current plan 
does not represent the political will of Georgia’s citi­
zens. See Part I, supra. The issue then became 
where we should look to divine Georgia’s true politi­
cal will. Given the unique events of Georgia’s 1992 
redistricting process, the Court has a better under­
standing of what the legislature might have done had 
it not been for the DOJ’s subversion of the redistrict­
ing process.



13a

We drew upon Georgia’s 1972 plan and 1982 plan, 
as well as the first plan enacted in 1992,® to deter­
mine how the legislature typically maintains district 
cores from one plan to another. Our first hurdle 
was where to locate the new Eleventh District. Since 
significant population growth is occurring in the At­
lanta area, it follows that the new district would be 
created where future growth is anticipated. Carving 
the new district where growth is projected allowed 
us to include fewer counties within the district, con­
sequently enabling us to better maintain district 
cores.

We placed the new Eleventh District in the North­
east Atlanta corridor out to the northeast Georgia 
state line. The Eleventh district now contains coun­
ties which are becoming more urban with the pop­
ulation growth, notably Walton County and Newton 
County. Eleventh Trl. Trans. I l l  at 26. The Elev­
enth also contains the Athens-Clarke County area 
which has become a metropolitan area in its own 
right, and locating it in the Eleventh is consistent 
with the new Eleventh’s urban/suburban flavor. Id.

The resulting Eleventh District is a relatively com­
pact grouping of counties which follow a suburban 
to rural progression and have Interstate Eighty-Five 
as a very real connecting cable. The road net, the 
area’s commerce, its recreational aspect, and other 
features produce a district with a palpable commu-

® The first 1992 plan gave us a clearer understanding of 
legislative intent, but it was not a perfect guide since it con­
tained many of the same features in its Eleventh District 
that found their way into the unconstitutional Eleventh— 
connecting the south Dekalb County urban minority popula­
tion with the primarily rural east Georgia minority popula­
tion.



14a

nity of interests. As a “radius” district reaching 
from suburban Atlanta to the state line, the new 
Eleventh has an analogous resemblance to the Third, 
Seventh, and Ninth Districts.

Georgia’s Tenth District has historically been lo­
cated in the east-central portion of the state, with 
Augusta as its main hub. Eleventh T rl Trans, III 
at 26; 1940 Plan, Jt. Exh. 16; 1972 plan. Addition 
of the Eleventh precipitated a shift in the Tenth 
District south, but the Tenth retains Augusta as its 
main hub and retains its primary rural character. 
The Court began here, and the remainder of the 
remedy fell into place along the lines of Georgia’s 
historic plans, maintaining district cores where pos­
sible.

2. Georgia’s Traditional Districting Factors
While Georgia’s legislature considers many factors 

in the apportionment process, there are a few fac­
tors that have historically guided the legislature in 
the process more so than others and which are 
evident in the plans it has historically enacted.'^ 
These factors, all of which have heavily influenced 
past apportionment plans in Georgia, are the pri­
mary factors that guided our development of this

Other factors the legislature has considered include avoid­
ing contests between incumbents, drawing districts to include 
certain state political officers, and drawing districts so as to 
avoid having certain state officers represented by members 
of the opposite political party. Eleventh Trl. Trans. I ll at 21. 
Because these were not predominant factors the state has 
considered in the past, they were given less weight since they 
are inherently more political in nature than the more objective 
factors.



15a

remedy. We explore some of them below, and we 
note that consideration of them was tempered by the 
one person-one vote principle.

a. Maintaining Political Subdivisions
The Georgia General Assembly has strongly ad­

hered to the principle of maintaining counties wihin 
the same congressional district. Georgia did not split 
a county until 1972, when it was necessary to do so 
because of the dense population in the metropolitan 
Atlanta area. See 1972 Map. The 1982 plan only 
split three counties, again in the Atlanta area for 
population density reasons.'^ Eleventh Trl. Trans. I ll 
at 36. Likewise, our remedy splits only six counties, 
all within the Atlanta area, and all split for popula­
tion density reasons.^ In comparison, the current 
unconstitutional plan split twenty-three counties, 
seventeen of which were split for no reason other 
than race.

Maintaining Dekalb County entirely within one 
district was a feature of several plans submitted to 
the Court. However, we chose to split Dekalb, remov­
ing a few voting precincts from the Fourth District 
and placing them in the Fifth District in order to 
achieve near perfect deviations in those Districts.*®

As noted above, the large number of counties in Georgia 
affords a draftsman much flexibility in achieving near per­
fect deviation numbers while maintaining county boundary 
lines.

Evidence showed that the significant population growth 
in Georgia is occurring in the metropolitan Atlanta area. 
Given the population density of those counties, it would be 
impossible to avoid splitting any counties.

Since Dekalb County is a densely populated county in the 
Atlanta area, splitting it is acceptable under the one person-



16a

The other reasons for doing this were: 1) the Geor­
gia legislature has already made the political decision 
to split Dekalb for population density reasons, first 
in 1982, and again in each of the three plans pre­
sented to the DOJ for preclearance in 1992; 2) the 
result was an increase in the Fifth District’s mi­
nority voting age population from fifty-three percent 
to fifty-seven percent, an increase most likely re­
quired by the VEA; and 3) the City of Atlanta, 
a significant political subdivision in its own right, is 
encompassed completely by the Fifth District.

b. Four “Comer Districts”
Georgia historically has maintained four “corner 

districts” in its Congressional plans. Eleventh Trl. 
Trans. I ll  at 23-27. The four “corner” districts in­
clude : 1) the southeastern coastal district, compris­
ing the coastal counties and the other counties most 
closely related to them; 2) an agrarian district in 
southwest Georgia, prominent for its peanut produc­
tion; 3) the northwestern corner, known for its car­
pet production and isolated from the northeast Geor­
gia counties by mountains; and 4) the northeastern 
corner, identified by its preeminent position in poultry 
production. Maintaining these districts is consistent 
with the community of interest factor approved by

one vote principle. The legislature split it in the past to com­
ply with that principle.

See Part 11(B) (2) (f) (1) (hi) (discussing the impor­
tance of maintaining black percentage of registered voters 
as close to fifty-five percent as possible). With the exception 
of four Dekalb precincts moved to the Fifth District, the rest 
of the Dekalb precincts were part of the Fifth in the 1982 
plan and remain in the Fifth under our plan.



17a

the Supreme Court. Johnson II, at 782. Our remedy 
adheres to all of these objectives with the exception 
of the northeast corner district, a necessary conse­
quence of adding an eleventh district and the shift in 
counties that it precipitated.

c. Urban Minority District
In 1972 Georgia created an urban minority dis­

trict in the Atlanta area to comply with the VRA. 
This is Georgia’s sole majority-minority congressional 
district. Recognition of the Fifth District as an ur­
ban minority district resulted from voting rights liti­
gation in the early seventies. Eleventh Trl. Trans. 
I ll  at 27. Since race is a factor this Court can con­
sider, we considered it insofar as the legislature 
would have considered it in maintaining one majority- 
minority district or else run afoul of VRA Sections 
2 and 5.

Unlike the unconstitutional Eleventh District, 
Fifth District residents have a strong community of 
interest, and the geography of the district itself is 
compact and meets contiguity requirements. See Jt. 
Exh. 15; Jt. Exh. 1; Remedy, Appendix A. Also, Sec­
tion 2 of the VRA required the Court to maintain the 
Fifth District as a majority-minority district. See 
Part 11(B) (2) (f) (1), infra.

d. Maintaining District Cores
Georgia’s legislature has historically sought to 

maintain district cores from one plan to another. The 
1972, 1982, and the three 1992 plans that passed the 
legislature evidence this legislative preference. Our 
remedy bears a striking resemblance to these earlier 
plans, especially in the rural districts such as the



18a

First, Second, Eighth, and Ninth. See Part 11(A), 
supra (noting that the remedy maintains ninety-five 
counties in the same districts as they were located 
in the 1982 plan, and the new Eleventh District con­
tains thirteen of the remaining counties).

e. Protecting Incumbents
Georgia has historically sought to draw district 

lines so incumbents remain in their districts in a new 
plan and to avoid placing two incumbents in the same 
district. The Court considered this factor, but since 
it is inherently more political than factors such as 
communities of interest and compactness, we sub­
ordinated it to the other considerations. We note, 
however, that our remedy places incumbents in the 
same district in only twO’ districts, while keeping eight 
incumbents in their present districts. The remedy’s 
Eighth and Eleventh Districts do not currently have 
incumbents as residents. The incumbent from the 
former Eighth District resides in the remedy’s Second 
District, and the Eleventh District incumbent resides 
in the remedy’s Fourth District.

f. Voting Rights Act 
1. Section 2

Section 2 of the Voting Rights Act, 42 U.S.C. 
§ 1973, applies to any “voting qualification or pre­
requisite to voting or standard, practice, or procedure 
. . . imposed or applied by any State or political sub­
division ----- ” 42 U.S.C. § 1973(a). A plain reading
of the statute would lead one to believe that Section 
2 does not apply to a court-drawn and ordered redis­
tricting plan. See Holder v. Hall, -----  U.S. ____•



19a

129 L.Ed.2d 687 (1994) (Thomas, J. Dissenting). 
However, Section 2 does apply to certain cases in 
which a federal court orders a redistricting plan into 
effect. See United States v. Dallas Cnty. Comm’n, 
850 F.2d 1433 (11th Cir. 1988) (holding that any 
proposed remedy of a Section 2 violation must meet 
Section 2 requirements), cert, denied, 490 U.S. 1030 
(1989).

During the remedy phase of the hearing, a signifi­
cant portion of the argument centered on whether 
the Court’s remedy would be required by Section 2 to 
contain two majority-minority districts. We have con­
sidered Section 2 and the case law construing it, and 
we conclude that two majority-minority districts are 
not required by Section 2 because including two such 
districts would violate Johnson IPs principles.

i) When Section 2 Requires Creation 
of a Majority-Minority District

States run afoul of Section 2 when they draw dis­
trict lines which have the “effect of denying a pro­
tected class the equal opportunity to elect its candidate
of choice.” Voinovich v. Quilter,------U.S. ------ , 122
L.Ed.2d 500, 512 (1993) (emphasis in original). 
Section 2 mandates that courts look at the totality of 
the circumstances when determining whether a viola­
tion exists. 42 U.S.C. § 1973(b). Plaintiffs establish 
a Section 2 violation by proving three factors, as first 
stated in Thornburg v. Gingles, 478 U.S. 30 (1986); 
(1) that the minority group “ ‘is sufficiently large 
and geographically compact to constitute a majority 
in a single-member district;’ . . .  (2) that the minor­
ity group ‘is politically cohesive; . . . [and] (3) that 
the white majority votes sufficiently as a bloc to en­
able it . . . usually to defeat the minority’s preferred



20 a

candidate.’ ” Voinovich, 122 L.Ed.2d at 513-14 (quot­
ing Growe, 122 L.Ed,2d at 403). However, these 
factors are not to be applied mechanically, Johnson
V. DeGrandy, -----  U.S. --------; 114 S.Ct. 2647, 2655
(1994); Voinovich, 122 L.Ed.2d at 514, and while 
they are the preconditions a plaintiff must first estab­
lish to prove a Section 2 violation, they must be con­
sidered along with other factors when analyzing a 
Section 2 claim. DeGrandy, 114 S.Ct. at 2657.

ii) Why Section 2 Does Not Require 
a Second Majority-Minority Dis­
trict in Georgia

Ample evidence has been presented during the two 
trials to enable the Court to determine that Section 2 
does not require creation of a second majority- 
minority district in Georgia—due primarily to the 
geographic dispersion of its minority population and 
lack of any significant vote polarization. Creating a 
second majority-minority district would also violate 
the Supreme Court mandate under which we now 
operate because Johnson II held that the VRA cannot 
be applied to reach an unconstitutional result.

Analysis of a racial map of Georgia reveals that 
the state’s minority population is widely dispersed.^

16 There are heavy minority concentrations in south Dekalb 
and Fulton counties, east central Georgia (most of which is in 
the new Eleventh District), southeast Georgia in the coastal 
region (all of which are included in the new First District), 
the southwest corner of Georgia, and a concentration in the 
west-central portion of the state. PI. Exhs. 17-23, 36-40. Cf. 
Def. Exh. 17; U.S. Exhs. 8-16. The only way Georgia could 
draw a majority-minority district in the southwest corner of 
the state was to use “narrow land bridges . . . and . . .  a 
number of irregular appendages.” Johnson III, at 6. The



21a
In fashioning a remedy, we considered the possibility 
of creating a second maj ority-minority district and 
concluded that to do so would require us to subordi­
nate Georgia’s traditional districting policies and 
consider race predominantly, to the exclusion of both 
constitutional norms and common sense. Such a plan 
would directly contravene the Supreme Court’s hold­
ing in Johnson II—that race cannot predominate in 
the redistricting context absent a compelling justifi­
cation for it. 132 L.Ed.2d at 782. Nonracial factors 
do not support creation of a second majority-minority 
district largely because Georgia’s minority population 
is not geographically compact. DeGrandy, 114 S.Ct. 
at 2655. Creating a second majority-minority dis­
trict would require this Court to engage in the uncon­
stitutional racial gerrymandering characteristic of 
the plan we now replace. Georgians deseiwe a better 
fate. “It takes a shortsighted and unauthorized view 
of the Voting Rights Act to invoke that statute . . . 
to demand the very racial stereotyping the Four­
teenth Amendment forbids.” Johnson II at 787.“

only way Georgia could create a maj ority-minority district 
out of the minority concentrations in east-central Georgia 
was to link that rural minority population, again using land 
bridges and appendages, to the large urban minority popula­
tion in south Dekalb County. See 1992 Plan. We will not 
repeat the same mistake.

Georgia could consider race as a predominant factor in 
redistricting if such a consideration was necessary to meet 
strict scrutiny. Johnson II, 132 L.Ed.2d at 782. We cannot 
subvert Georgia’s traditional districting principles in order 
to draw a second majority-minority district. If Georgia had 
a concentrated minority population large enough to create 
a second majority-minority district without subverting tra­
ditional districting principles, the Court would have included 
one since Georgia’s legislature probably would have done so. 
The VRA does not require proportional representation, how-



'Ifl.dt,

Statistical evidence pertaining to statewide voting 
patterns in Georgia showed a significant degree of 
crossover voting. Johnson I, 864 F. Supp. at 1390.

ever. See, e.g., White v. Regester, 412 U.S. 755, 765-66 
(1973); Whitcomb v. Chavis, 403 U.S. 124, 148-56 (1971). 
As the Supreme Court stated in Shaw v. Reno:

The principle of equality is at war with the notion 
that District A must be represented by a Negro, as it is 
with the notion that District B must be represented by 
a Caucasian, District C by a Jew, District D by a Catho­
lic, and so on. . . . That system, by whatever name it is 
called, is a divisive force in the community, emphasizing 
differences between candidates and voters that are irrele­
vant in the constitutional sense.

125 L.Ed.2d 511, 529 (1993) (quoting Justice Douglas’ dis­
sent in Wright v. Rockefeller, 376 U.S. 52, 66-67 (1964)).

At the remedy phase of the October, 1995 hearing, the 
counsel for the Speaker of Georgia’s House of Representatives 
observed that if the Court did include a second majority- 
minority district, it would be “set in stone.” He contended 
that, since future plans would be measured against the Court’s 
plan for purposes of VRA analysis, Georgia’s legislature 
would be hamstrung to relocate the district elsewhere, or to 
alter the mix of counties included within the district, for 
fear of violating the VRA. Counsel for Amicus Curiae agreed, 
which is significant since the Amicus plan contains two 
majority-minority districts. Cf. Johnson I, 864 F. Supp. at 
1386 (noting problems a plan attempting to give minorities 
proportional representation could present in drafting future 
plans to satisfy VRA). Since political considerations pervade 
the redistricting task, the Court feels that any permanent 
footprint left on Georgia’s political landscape, especially one 
with such dangerous side effects as the Supreme Court has 
noted, should be left to those elected to make such decisions. 
In other words, we feel the need to intrude as little as possi­
ble in this inherently political territory.



23 a

We observed that “ [b]lack and black-preferred candi­
dates in Georgia have achieved many electoral vic­
tories in local and statewide elections and have 
received significant—occasionally overwhelming—sup­
port from both black and white voters within the 
Eleventh Congressional District/ Id. at 1390-91. 
During the Second District trial, Linda Meggers 
stated that election results in that district indicated 
white crossover voting. Second Trl. Trans, at 438. 
Indeed, Congressman Sanford Bishop, the black Con­
gressman currently representing the Second District, 
agreed. Id. at 142. There is conflicting evidence on 
the degree of vote polarization in Georgia. This 
much is clear: while some degree of vote polariza­
tion exists, it is “not in alarming quantities.” 
Johnson I, 864 F. Supp. at 1390. The second and 
third Gingles preconditions therefore would not be 
met in a Section 2 challenge to the remedy. Thus, 
the remedy we now impose meets the requirements of 
Section 2 without containing two majority-minority 
districts.

We note that several districts in the remedy do 
contain significant minority populations which could 
influence election results. The 1982 plan contained 
five districts with a minority population of greater 
than twenty-five percent, and three with a minority 
population of greater than thirty-five percent. Our 
remedy contains six districts with a minority voting 
age population of greater than twenty-five percent, 
and two districts, the Fifth and the Second, with a 
minority voting age population of greater than thirty- 
five percent, while the remedy’s Tenth comes close to 
that with a minority voting age population of more 
than thirty-four percent. The remedy’s minority 
percentages per district are comparable to those in



24a

the 1982 plan, even with the addition of the Eleventh 
District.

While we were required to use the 1990 census 
numbers in formulating the remedy, we note the 
demographic trend indicating that Dekalb County’s 
population, which comprises the majority of the rem­
edy’s Fourth District, is becoming more heavily 
minority. This means that the Fourth District has 
the potential to become a very strong minority influ­
ence district or a majority-minority district as Geor­
gia is reapportioned following future censuses, if it 
is not already.”

iii) What Section 2 Does Require
While Section 2 does not require the Court to cre­

ate a second majority-minority district, it does re-

In its response to the Court’s request for additional 
demographic information concerning minority population 
trends in Dekalb County, the State Defendants produced an 
affidavit of Linda Meggers. Ms. Meggers attests that while 
in 1990 27.4% of Dekalb County’s registered voters were 
black, that number had increased to 35.7% by 1994. Meg­
gers’ 11/21/95 Aff. 11 7. She indicated the number could be 
as high as 39.4% in 1995, but that that number is due to a new 
voter registration law more than population shifts. Id. at 
Ifll 7-9. In any event, while the Court is bound to use 1990 
census figures in formulating a remedy, we cannot ignore 
the strong trends in Dekalb County which indicate the rem­
edy’s Fourth District may be a stronger minority influence 
district than data based on the 1990 census would indicate.

The Abrams Interveners do not dispute Meggers’ figures; 
they only caution the Court that those figures must be viewed 
in light of the overall change in percentages for each dis­
trict because some may have lost minority population. Their 
warning is duty noted.



25a

quire maintenance of the Fifth District as a majority- 
minority district. The minority population in the 
Fifth District is sufficiently compact and, being an 
urban minority population, has a sufficiently strong 
community of interest to warrant being a majority- 
minority district. Compliance with the VRA required 
creating the Fifth as majority-minority in 1972. 
Eleventh Trl. Trans. I ll  at 27. As a result of 1982 
litigation, the Fifth District was required to have a 
minority population of sixty-five percent. Busbee v. 
Smith, 549 F. Supp. 494 (D.D.C. 1982), aff’d, 459 
U.S. 1166 (1983).

The minority population of sixty-one percent, 
minority voting age population of fifty-seven percent, 
and a black registered voter percentage of fifty-four 
percent in the remedy’s Fifth District does not result 
in a dilution of the rights of that district’s minori­
ties. Dr. Katz, the most credible statistical expert at 
the first trial, identified a ‘̂significant majority” 
minority district as one having fifty-five to sixty per­
cent of black registered voters. See Katz Rep. at 22. 
We endeavored to maintain the black percentage of 
registered voters in the Fifth District as close to 
fifty-five percent as possible. Expert testimony at the 
Eleventh District trial indicated a significant drop 
in the probability of a black candidate being elected 
as the black percentage of registered voters declines 
to fifty percent. Katz Rep. at tab 5. The probability 
of electing a black candidate is below fifty percent 
when the percentage of black registered voters equals 
fifty percent. Id. Therefore, maintaining the per­
centage of black registered voters as close to fifty-five 
percent as possible was necessary, in our view, to



26a

avoid dilution of the Fifth District minorities’ 
rights.'̂ '®

The minority population of the Fifth District in 
the 1982 plan was sixty-five percent in order for 
Georgia to comply with the VRA. Bushee, supra. 
We felt that allowing the minority population to fall 
below sixty percent might be viewed as dilutive at 
some level. The minority population in the remedy’s 
Fifth District is just shy of sixty-two percent.“

2) Section 5
VRA Section 5 requires certain states which have 

a history of discriminating against minorities to seek 
preclearance from the U.S. Attorney General prior 
to implementing a new districting or apportionment 
plan. 42 U.S.C. § 1973c. However, [a] decree of the 
United States District Court is not within reach of 
Section 5 of the Voting Rights Act.” Connor v. John­
son, 402 U.S. 690, 691 (1971); McDaniel v. Sanchez, 
452 U.S. 130, 138 (1981). Thus, this Court is not 
required to seek the preclearance of the Attorney 
General prior to implementing its plan. The Court 
does note that its remedy is not retrogressive to the

18 While we noted flaws in these analyses in Johnson I, these 
provide the most probative analyses we have which indicate 
what might be dilutive in the Fifth District. These statistics 
are based on statewide figures, though. The Court hesitates 
to rely on statistical data using percentages of black registered  
voters as its basis since that in essence condones voter apathy.

We believe minority voting age population is the appro>- 
priate measure for dilution analysis. We offer minority pop­
ulation percentages strictly for comparison purposes since 
that is what the Bushee court focused on in essentially forcing 
Georgia to create a Fifth District in 1982 with a minority 
population of 65%.



27a

rights of minorities when compared to the 1982 plan,®® 
and it does not violate Section 5 for the same primary 
reasons that it does not violate Section 2.

Connor notwithstanding, some court-ordered plans 
have been subject to Section 5 and its requirements. 
In Sanchez, the Supreme Court held that Section 5’s 
preclearance requirement did apply to a reapportion­
ment plan submitted to a federal court by the legis­
lative body of a jurisdiction covered by Section 5. 452 
U.S. at 153.

It could be argued that the Court’s remedy is a 
legislative plan since it resembles some of the legis­
lative plans, including the unconstitutional one. To 
varying extents we relied upon the three enacted 1992 
plans, as well as the 1972 and 1982 plans, to deter­
mine how the legislature maintains district cores and 
communities of interest, so some similarity to those 
plans is inevitable.

However, the Court’s plan is its own. While it may 
resemble one or more of several plans presented to 
us, as a member of this Court observed during the 
remedy phase of the hearings: “There could come a 
time when even the fertility of Herschel the [reappor- 
tionment] computer is exhausted and we have re-

Debate arose at the August, 1995 hearing and the Oc­
tober, 1995 trial as to whether the current plan or the 1982 
plan is the appropriate benchmark for retrogression analysis. 
Interestingly, the DOJ asserted that the unconstitutional plan 
should be used, contravening 28 C.F.E. § 51.54(b) (1) which 
dictates using the last constitutional plan on the ground if 
the current plan is unconstitutional. Since the Court is in the 
position of replicating the legislative task as it existed in 
1992, the 1982 plan should serve as a guide. Of course, since 
the Court’s plan is not subject to Section 5 preclearance, the 
point is academic.



28a

ceived a finite number of maps. . . .  If we extend that 
logic we are going to be rubber stamping somebody’s 
plan, but I suspect that what we’ll do will be our own 
unique product.” Second Trl. Trans, at 409.

So the remedy we now order is not the product of 
Georgia’s legislature. We of course would have pre­
ferred for the legislature to have fulfilled its obliga­
tion and formulated a districting plan for Georgia. 
Alas, none was presented to us.®̂  We therefore have 
formulated our remedy in adherence to constitutional 
norms and deference to the legislature’s historical 
preferences. Therefore, it is not subject to preclear­
ance under VRA Section 5.

III. Conclusion
The Court has fashioned a plan to remedy Georgia’s 

current, unconstitutional districting plan. In arriv­
ing at this remedy, we adhered as closely as possible 
to the Supreme Court’s mandate in Johnson II, the 
principle of one person-one vote, and Georgia’s tradi­
tional legislative goals and traditions. Adopting a 
remedy that would be truly minimally disruptive to 
Georgia’s current plan was not an option. No plan of 
congressional redistricting created by court or legis­
lature will achieve perfection. There are too many 
practical, political, and altogether human features in 
the equation. We do no harm with this plan, which 
cures the unconstitutionality of the former and can 
serve in “caretaker” status until the legislature con-

Some members of the General Assembly did present plans 
to the Court, but none purported to be the actual product of 
the General Assembly. Nevertheless we considered them along 
with the other plans presented to us during this process.



29 a

venes tO' change it. That may occur following the mil­
lennium census, or before.

Noting again the reluctance with which the task 
was undertaken, it is hereby ORDERED and DE­
CREED that henceforth all elections for the members 
of the House of Representatives of the Congress of 
the United States from the State of Georgia shall be 
conducted in accordance with the plan which is ap­
pended hereto as Appendices “A” and “B.”

SO ORDERED, this 13th day of December, 1995.

/ s /  B. Avant Edenfield 
B. Avant E denfield 
Chief Judge
United States District Court 
Southern District of Georgia

/ s /  Dudley H. Bowen, Jr.
D udley H. B ow en, Jr.
United States District Court 
Southern District of Georgia

EDMONDSON, Circuit Judge, dissents.

/ s /  JLE

[Appendices Omitted]



30a

APPENDIX B

[Filed Dec. 1,1995]

IN THE UNITED STATES DISTRICT COURT 
FOR THE SOUTHERN DISTRICT OF GEORGIA 

AUGUSTA DIVISION

No. CV 194-008

Davida J o h n so n ; Pam  B u r k e;
Henry  Zittrouer; George L. D eloach ; 

and George Seaton, plaintiffs

vs.

Zell Miller, in his Official Capacity as Governor of 
the State of Georgia; P ierre H oward, in his Of­
ficial Capacity as Lieutenant Governor of the State 
of Georgia and President of the Georgia Senate; 
T homas Murphy , in his Official Capacity as 
Speaker of the Georgia House of Representatives 
and Max Cleland , in his Official Capacity as Sec­
retary of State of the State of Georgia, defend- 
ANTS

Lucious A brams, Jr .; Rev. G. L. Avery; W illiam 
Gary Chambers, Sr.; Judy Lam bers; R ita Va­
lenti ; and Karen W atson , intervenor-defend-
ANTS and

U nited States of A merica, 
intervenor defendant



31a

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

ORDER
In this expedited matter, the question presented is 

this one: Is Georgia’s Second Congressional District 
unconstitutional on the ground that it violates the 
Equal Protection Clause of the Fourteenth Amend­
ment to the United States Constitution? The answer 
is “Yes.”

After a trial on the merits and in light of Miller 
V. Johnson, 115 S.Ct. 2475 (1995), we find and con­
clude, that race, namely the intention to create a 
congressional district in which black persons would 
be a majority of the voting age population, was the 
overriding and predominant factor motivating the 
placement of the Second District’s boundaries.^ We 
find that the boundary line for the district was con­
sistently drawn to keep potential black voters in the 
district and to keep potential white voters out of the 
district. The district was drawn to segregate voters 
according to their race; so, it is an unlawful district.

1 Some of the parties contended that no trial was necessary. 
The Department of Justice contended that, as a matter of 
law, the Second District was unconstitutional considering this 
court’s earlier judgment and the Supreme Court’s affirmance 
of that judgment. Plaintiffs contended that the pretrial record 
in this case entitled them to summary judgment as a matter of 
law. Out of an abundance of caution and given the need to 
move forward as conveniently and speedily as the court could, 
a trial was held; and the legal issues raised by the Depart­
ment of Justice and the Plaintiffs on the question of whether 
a trial was needed were bypassed.



32a

Findings of Fact
The facts stipulated to by all of the parties are 

accepted and found by the court. The court also 
stands by the findings of fact in its earlier decision 
on Georgia’s Eleventh District, Johnson v. Miller, 
864 F.Supp. 1354 (S.D. Ga. 1994);'^ those findings 
are pertinent to the Second District as well.

The first congressional redistricting plan submitted 
by Georgia to the United States Attorney General 
for preclearance in October 1991 contained a Second 
Congressional District, located in southwestern Geor­
gia, in which blacks comprised 35.37% of the voting 
age population (“VAP” ). The Department of Jus- 
tive (“DOJ”) refused preclearance of this plan. A 
new plan was enacted and submitted for preclearance 
in which the black VAP in the Second Congressional 
District was 45.01%. The DOJ again refused pre­
clearance, relying on alternative plans proposing 
three majority-minority districts.

The critical element to making the Second Con­
gressional District a majority-minority district was 
the “Macon/Savannah trade” which transferred black 
voters from Macon—located in the Eleventh Con­
gressional District in the first two state plans sub­
mitted to the DOJ—to the Second Congressional Dis­
trict and then extending the Eleventh District into 
Savannah. This move converted the Second Con­
gressional District into a majority-minority district

2 The record of the trial on the Eleventh District was made 
part of the record for the trial on the Second. At the trial of 
the Second District, many witnesses testified either in person 
or by deposition; and many documents were introduced. All 
the evidence has been considered, although not all of it will be 
specifically mentioned in this opinion.



33a

and kept the Eleventh Congressional D istrict a 
majority-minority district by offsetting the loss of 
black voters in Macon with the addition of black 
voters in Savannah. This third plan was precleared 
by the DOJ. As enacted, the Second Congressional 
D istrict had a 52.33% black VAP.

Twelve of the Second Congressional D istrict’s 
thirty-five counties are split, and to draw the Second 
Congressional D istrict as it is now also required the 
splitting of twenty-eight precincts.® In  addition, the 
boundary of the Second District splits the cities of 
Columbus, Macon, Albany, Manchester, Roberta, By­
ron, Centerville, W arner Robins, Ft. Valley, Perry, 
Cordele, Leesburg, Moultrie, and Valdosta. Before 
1992, the Second District included no portions of the 
city of Columbus or portions of the City of Macon. 
Before 1992, the Second Congressional D istrict con­
tained no split counties.

We find that the sole reason for splitting precincts 
was racial and tha t the predominant reason for split­
ting the counties and cities was racial as well.

Georgia’s Second District makes use of narrow 
land bridges to connect parts of the district and in-

® By the way, one bad side effect of the splitting of units 
as small as precincts for racial purposes occurs when the 
precinct is divided in one way for a state house seat, another 
way for a state senate seat, and yet a different way for a con­
gressional seat. When this kind of gerrymandering is re­
peated often enough, the voting combinations that appear 
on some ballots are so rare—at the intersection of the state 
house, state senate, and congressional district lines within a 
precinct—-that a secret ballot becomes difficult for some voters. 
Given the small number of ballots showing the specific com­
bination of candidates, the persons later counting the few 
ballots of that kind actually cast might well know which ballot 
a particular voter cast.



34a
volves a number of irregular appendages. These fea­
tures affect the district’s compactness adversely and, 
in some instances, make parts of the district only 
barely contiguous. We find tha t the predominant 
reason for these irregular lines is race; most can be 
explained on no other basis.

Linda Meggers, Director of Reapportionment Serv­
ices for the Georgia General Assembly, was qualified 
as an expert witness. She testified that it was not 
feasible to create a majority-minority district in the 
Second Congressional D istrict without including the 
black population centers in Columbus and Muscogee 
County, Albany and Dougherty County, and Macon 
and Bibb County. She fu rther testified that, in draw­
ing the Second Congressional District, she—we find 
her to have been the chief draftsperson for the Dis­
tric t—followed the “Max-Black Plan,” as espoused by 
the ACLU, to the same degree to which she followed 
it in drawing the Eleventh Congressional District. 
The Max-Black Plan’s purpose was to maximize black 
voting strength in certain congressional districts by 
making the racial composition of those districts the 
overriding consideration in their design. She also 
said that, in terms of socio-economic interests in­
cluded in the District, the Second District is one of 
the most diverse in Georgia. We credit Ms. Meggers’ 
testimony.

A comparison of maps depicting the Second Dis­
tric t’s twelve split counties with maps showing the 
concentration of black residents in the same counties 
proves that the drawing of Georgia’s Second Con­
gressional D istrict was motivated predominately by 
racial considerations.^ P u t differently, the line was

* The portions of Bibb County placed in the Second District 
reflect the heavy concentration of black residents in that por-



35a

drawn to put black voters in the Second District and 
to keep white voters out.

Dr. Timothy O’Rourke, Professor of Citizenship 
Education a t the University of Missouri-St. Louis, 
was qualified as an expert witness. Dr. O’Rourke tes­
tified that he had been asked to examine the bound­
aries of the Second Congressional D istrict in the light 
of the standard announced by the Supreme Court in 
Miller v. Johnson and to form an opinion about 
whether the Georgia Legislature had, in fact, sub­
ordinated its traditional redistricting principles to 
race for the Second District. He fu rther testified tha t 
he had taken into account such m atters as the one- 
person/one-vote standard, respect for political sub­
division boundaries, compactness, contiguity, the ex­
ten t to which the district includes metropolitan areas, 
the extent to which the district includes media m ar­
kets, and socio-economic communities of interest in the 
region. By following a process of elimination which 
he explained. Dr. O’Rourke concluded th a t the pre­
dominant motivating factor for the configuration of

tion of Bibb County. Pltff.Exh. 1, 13 (Joint Exh. 2, Pltf. 
Exh. 32). The portions of Colquitt County placed in the 
Second District reflect the heavy concentration of black resi­
dents in that portion of Colquitt County. Pltff.Exh. 2, 14 
(Joint Exh. 9, Pltf.Exh. 38). The portions of Crisp County 
placed in the Second District reflect the heavy concentration 
of black residents in that portion of Crisp County. Pltf.Exh. 
3, 15 (Joint Exh. 8, Pltf.Exh. 37). The same can be said for 
the remaining nine counties that are split in the Second Con­
gressional District. Pltf.Exh. 4-12, 16-23 (Joint Exh. 3-7 
and 10-13, Pltf.Exh. 33-36 and 39-42).

To alleviate confusion, we cite to both sets of exhibits— 
those submitted as part of Plaintiffs’ motion for summary 
judgment and those used at trial.



36a

the Second District was race. We credit Dr. 
O’Rourke’s testimony.

We find tha t Georgia did discriminate against its 
black citizens in voting m atters in the past.

We find no evidence that Georgia’s General As­
sembly a t the time the Second Congressional D istrict 
was draw n was, in reality, seeking to eliminate the 
effects of past racial discrimination on black voting 
patterns in southwest Georgia.

We find tha t the General Assembly’s intent a t the 
time the Second Congressional D istrict was drawn 
was to comply with the Justice Department’s in ter­
pretation of the Voting Rights Act and the Depart­
ment’s preclearance demands.

Conclusions of Law
In defense, the State defendants (“State” ) stress 

only tha t the Second District differs from the uncon­
stitutional Eleventh Congressional D istrict because 
(1) the Second D istrict has always existed in the 
southwestern comer of the State, (2) there is a 
greater community of interest in the Second District, 
and (3) the Second District, compared to the Eleventh 
District, has a lower percentage of black voters. 
While it is trae  th a t the Second District does differ 
in some respects from the Eleventh, the differences 
are not significant enough to make the Second District 
constitutional.

Nothing in the State’s argum ent changes the fact 
tha t the General Assembly was predominantly moti­
vated by race in its drawing of the Second Congres­
sional District. The sections of the district’s bound­
ary that were drawn for racial reasons are substan­
tial in terms of both distance and population affected.



37a

The voters within the Second D istrict may possibly 
have a greater community of interest than those in 
the Eleventh District, but it  is clear the General As­
sembly did not draw many of the lines of the Second 
D istrict because of that community of interest. Look­
ing a t the split counties and precincts, white voters 
left out of the district because they were white shared 
the same interests (for example, agriculture or mili­
ta ry  employment) as black voters who were included 
because they were black. That fewer black voters 
were purposely placed into the Second D istrict than 
into the Eleventh does not change the fact tha t the 
General Assembly intentionally placed black voters 
into the Second and kept white voters out of the Sec­
ond solely because of their race.

Because race was the overriding and predominant 
motivating factor in designing the Second Congres­
sional District, the district can be upheld only if  it 
survives strict scrutiny. Miller, 115 S.Ct. a t 2490.

The State fails to meet its burden under the strict 
scrutiny analysis. The State never argued th a t the 
Second D istrict was required by the Voting Rights 
Act.® We accept th a t eradicating the effects of past 
racial discrimination can be a significant state in ter­
est. See Shaw v. Reno, 113 S.Ct. 2816, 2831 (1993). 
But, the State has not argued persuasively tha t the 
present Second Congressional D istrict was configured 
to eradicate the effects of past racial discrimination, 
as opposed to and apart from complying with federal

® The Supreme Court has said that the Second District— 
the third and last majority-minority district to be created 
for Georgia—was not required for Section 5 preclearance 
and that compliance with a mistaken reading of the Voting 
Rights Act cannot justify race-based districting. Miller, 115 
S.Ct. at 2491-92.



38a

statutes. See generally Miller, 115 S.Ct. a t 2490 (not 
every state action to comply with DOJ’s view of 
Voting Rights Act is, in itself, step by state to eradi­
cate effect of past racial discrimination for purposes 
of strict scrutiny). No evidence exists in the record 
th a t the General Assembly a t the time the Second 
D istrict was drawn was seeking, in reality, to elim­
inate the effects of past racial discrimination on black 
voting patterns in southwest Georgia. Cf. City of 
Richmond v. J.A . Croson Co., 109 S.Ct. 706, 719-721 
(1989) (discussing need for government to demon­
strate specific need and justification for favoring 
racial group).

We conclude and declare that Georgia’s Second 
Congressional D istrict is unconstitutional in its cur­
rent composition. Defendants are hereby barred from 
using it in fu ture Congressional elections.

IT IS SO ORDERED.

F or the  T hree-Judge Court :

/ s /  J. L. Edmondson
United States Circuit Judge



39a

APPENDIX C

[Filed July 26, 1994]

STATEMENT OF JUDICIAL NOTICE
Evidence of form er discriminatory practices against 

black people in the State of Georgia need not be pre­
sented for purposes of this case. Racial discrimina­
tion has been an unfortunate reality in the state’s 
history. No one can deny that state and local gov­
ernments of Georgia in the past utilized widespread, 
pervasive practices to segregate the races which had 
the effect of repressing black citizens, individually 
and as a group.

In the past in Georgia, like in so many other states 
of the Union, racial segregation was government pol­
icy and a way of life. By law, public schools and pub­
lic housing were segregated according to race. Public 
recreational facilities were segregated. Miscegenation 
was prohibited. Ordinances required segregation in 
public transportation, restaurants, hotels, restrooms, 
theaters, and other such facilities, even drinking 
fountains.

More indirectly, the government often treated black 
citizens differently from white citizens. Public serv­
ices were allocated along racial lines. To find roads 
in white neighborhoods paved but roads in black 
neighborhoods unpaved was common. In  public em­
ployment, black workers were often paid less than 
white workers for the same job. In addition, methods 
of ju ry  selection were developed to exclude black peo­
ple from jury  service.

Georgia’s history on voting rights includes discrim­
ination against black citizens. From the state’s first 
Constitution—which barred blacks from voting alto-



40 a

gether—through recent times, the state has employed 
various means of destroying or diluting black voting 
strength. For example, literacy tests (enacted as late 
as 1958) and property requirements were early 
means of excluding large numbers of blacks from the 
voting process. Also, white primaries unconstitution­
ally prevented blacks from voting in prim ary elec­
tions a t the state and county level.

Even after black citizens were provided access to 
voting, the state used various means to minimize 
their voting power. For example, until 1962 the 
county unit system was used to undermine the vot­
ing strength of counties with large black populations. 
Congressional districts have been drawn in the past 
to discriminate against black citizens by minimizing 
their voting potential. State plans discriminated by 
packing an excessive number of black citizens into a 
single district or splitting large and contiguous 
groups of black citizens between multiple districts.

Never do we imply that a history of invidious ra ­
cial discrimination is unique to Georgia. Our judicial 
notice is confined to Georgia because this case is 
about Georgia, and Georgia is a covered jurisdiction 
under the Voting Rights Act. Our notice is intended 
to set out the historical facts, not to insult a state or 
people who have made great and difficult strides to­
wards equality during the last th irty  years and who 
could be mentioned for many good, kind, and brave 
things unrelated to this case.



Court Plan
1995

41a

Library; Georgia 

Plan: LIT16620124FEDCT

m >
H »

Cm rrt y !Vxrvlat»ns

?6 an

Prepared by:
Vrdled Siaies DepartmenJ o f  Justice 
Washington, D C., 2 (530

February 23, /  995





43a

NOTE: VAP = Voting Age Population

District Total Pop Black Pop VAP Black VAP
Number % Deviation % of Total % of Total :% of VAP

1 588541 179809 426558 118631
-0 .0 7 30.55 72.48 27.81

2 587583 230419 415407 145948
-0 .2 3 39.21 70.70 35.13

3 589630 145377 424974 95713
0.12 24.66 72.07 22.52

4 589322 215700 450500 147784
0.07 36.60 76.44 32.80

5 589359 365330 442885 253212
0.07 61.99 75.15 57.17

6 589600 37597 440475 27122
0.11 6.38 74.71 6.16

7 589405 77871 432683 51911
0.08 13.21 73.41 12.00

8 587912 182636 425122 120107
-0 .1 7 31.07 72.31 28.25

9 589420 21520 439088 15164
0.08 3.65 74.49 3.45

10 588046 220803 423299 145950
-0 .1 5 87.55 71.98 34.48

11 589398 69503 429927 46600
0.08 11.79 72.94 10.84

Totals 6478216 1746565 4750913 1168142
Number of Districts 11
Members Per District 1
Ideal District Size : 588929
Average Deviation (%) : 0.11
Deviation Range (:%) : - 0.23 to 0.12
Overall Deviation (%) : 0.35

DATA SOURCE: 
NOTE:

1990 US Census PL94-171 Population Counts 
Districts numbered >  200 are used as special 
accumulators.
They are not included in avg or % range 
calculations.



44a

illustrative Plan Library: Georgia 

Plan: GAPLANA

■
■ 0

□ «
■ ’
B «

Cfxrrty lioimiflfxB

0  e.b 13 ?fi 30

Prepared by:
United States Departmeni o f  Justice 
Washington,D-C.,2(S30

February 2 0 ,1 996



45a
Plan: GAPLANA

Population Summary Report
Black

Dist Total Deviation Dev % Black \% 18 + %
1 588541 -3 8 7 -0 .0 7 30.55 27.81
2 588995 67 0.01 46.10 42.03
3 589121 193 0.03 18.48 16.92
4 588567 -3 6 1 -0 .0 6 10.21 9.54
5 588398 -5 3 0 -0 .0 9 58.21 53.70
6 588680 -2 4 8 -0 .0 4 6.48 6.23
7 589443 515 0.09 13.32 12.12
8 589202 274 0.05 27.95 24.88
9 589420 492 0.08 3.65 3.45

10 589476 548 0.09 27.10 24.47
11 588367 -5 6 1 -0 .1 0 54.60 51.04

Total Population 6478210
Ideal Population 588928
Mean Deviation is 0
Mean Percent Deviation is 0
Largest Positive Deviation is 548
Largest Negative Deviation is -5 6 1
Overall Range in Deviation is 1109
Overall Range in Deviation % is 0.19



46a

1991 Plan Library: Georgia

Plan: S913556MEW

□ 0 
a ^

H

Crt rrty fV»ndflnos

A
J L
I'Y

Scale in Miles

Prepared by:
IJruled Stales Departmeni o f  Justice 
Washington, D.C., 2(530

February 27.1996

0 6.b 13 ?fi



47a

APPENDIX D

IN THE UNITED STATES DISTRICT COURT 
FOR THE SOUTHERN DISTRICT OF GEORGIA 

AUGUSTA DIVISION

No. CV 194-008

Davida Johnson , et a l ., plaintiffs

V.

Zell Miller, et al ., defendants 

and
U nited States of A merica, 

defendant-intervenor

and
L ucious A brams, Jr ., et al ,, 

defendant-intervenors

UNITED STATES’ NOTICE OP APPEAL 
FROM DECEMBER 13, 1995 ORDER

Pursuant to 28 U.S.C. §§ 1253 and 2101, the 
United States hereby appeals to the Supreme Court 
of the United States from this Court’s order of De­
cember 13, 1995 adopting a remedial congressional 
redistricting plan for the State of Georgia and order­
ing that henceforth all elections for the House of 
Representatives of the Congress of the United States 
be conducted in accordance with tha t plan.



48a

Respectfully submitted,

H arry D. D ixon, Jr. Deval L .Patrick 
United States Attorney Assistant Attorney General

/ s /  Daniel H. Claman 
E lizabeth J ohnson 
D onna  M. Murphy  
Daniel  H. Clam an  
Attorneys, Voting Section 
Civil Rights Division 
Department of Justice 
P.O. Box 66128 
Washington, D.C. 20035 
(202) 514-6340

U . 5 .  SO V ER N M EN T P R IN T IN G  O P P lC B i ! 9 9 6 4 0 5 0 1 7  4 0 0 7 0