Abrams v. Johnson and United States v. Johnson Join Appendix
Public Court Documents
March 20, 1996
Cite this item
-
Brief Collection, LDF Court Filings. Abrams v. Johnson and United States v. Johnson Join Appendix, 1996. 5e831ec6-ab9a-ee11-be37-00224827e97b. LDF Archives, Thurgood Marshall Institute. https://ldfrecollection.org/archives/archives-search/archives-item/f336a34b-f32b-4289-a503-8f3b3fb72dfe/abrams-v-johnson-and-united-states-v-johnson-join-appendix. Accessed November 03, 2025.
Copied!
Nos. 95-1425 and 95-1460
3fn tf)£ ^upreim Court of tfjr Miutrb £§>tatr£
OCTOBER TERM, 1996
LUCIOUS ABRAMS, J R , ET A L , APPELLANTS
v.
DAVIDA JOHNSON, ET AL.
UNITED STATES OF AMERICA, APPELLANT
v.
DAVIDA JOHNSON, ET AL.
ON APPEAL FROM THE UNITED STATES DISTRICT COURT
FOR THE SOUTHERN DISTRICT OF GEORGIA
JOINT APPENDIX
David F. Walbert
Counsel o f Record
Walbert & Mathis
100 Peachtree Street
Suite 14.OO
Atlanta, GA 30303
(hOU) 523-5000
Michael J. Bowers
Attorney General o f
the State o f Georgia
(see inside cover for additional counsel)
Walter Dellinger
Acting Solicitor General
Department o f Justice
Washington, D.C. 20530
(202) 5U-2217
Jurisdictional Statements Filed:
March 7 and March 12,1996
Probable Jurisdiction Noted: May 20,1996
Dennis R. Dunn
132 State Judicial Building
Atlanta, GA 30334
Counsel fo r Appellees Miller et al.
A. Lee Parks
Counsel o f Record
Kirwan, Parks, Chesin & Remar, PC.
75 Fourteenth Street
2600 The Grand
Atlanta, GA 30309
(404) 873-8000
Counsel for Appellees Johnson, etal.
Laughlin McDonald
Counsel o f Record
Neil S. Bradley
Maha Zaki
Mary Wyckoff
American Civil Liberties
Union Foundation, Inc.D
44 Forsyth Street — Suite 202
Atlanta, GA 30303
(404) 523-2721
E laine R. J ones
Director-Counsel
Theodore M. Shaw
Norman J. Chachkin
J acqueline Berrien
NAACP Legal Defense and Educational Fund, Inc
99 Hudson Street
New York, N.Y. 10013
Gerald R. Weber
American Civil Liberties Union o f Georgia, Inc.
142 Mitchell Street, S.W.
Suite 301
Atlanta, GA 30303
(404) 523-6201
Counsel fo r Appellants Abrams, et al.
3fn tf)t B>u$mm Court of tfje MnitEb
OCTOBER TERM, 1996
No. 95-1425
LUCIOUS ABRAMS, JR., ET AL., APPELLANTS
v.
DAVIDA JOHNSON, ET AL.
No. 95-1460
UNITED STATES OF AMERICA, APPELLANT
v.
DAVIDA JOHNSON, ET AL.
ON A PP EA L FROM THE UNITED STATES
D ISTRICT COURT
FOR THE SOU THERN D ISTRICT OF GEORGIA
JOINT APPENDIX
TABLE OF CONTENTS
Page
A. Relevant docket entries in the United States
District Court for the Southern District
of Georgia, Augusta Division ...................... 1
From the 1995 Trial:
B. Testimony of Joseph K a tz .......... ................... 9
l
11
Page
C. [Corrected] Report of Joseph Katz, State’s
Exhibit 170 , with selected attachments from
Tabs 3 and 4 .................. ................... . 30
D. Testimony of Allan J. Lichtman ..................... 53
E. Report [by Allan J. Lichtman] on Issues
Relating to Georgia Congressional Districts,
Exhibit DOJ 24, with selected tables, and
charts from Appendix 1 ............................ 62
F. Testimony of Ronald E. Weber ...................... 84
G. Report of Ronald E. Weber, Plaintiffs’
Exhibit 82 and tables from Attachment E .. 96
H . Joint Statement of Undisputed Fact ............ 116
From the Remedial Phase:
I. Three declarations by Selwyn Carter
(Abrams Exhibit 35) ...................................... 141
J. Testimony of Linda M eggers.......................... 152
K. Testimony of Sanford Biship .......................... 155
L. Argument: Laughlin McDonald...................... 157
M. Testimony of Selwyn Carter ............... 159
N. Argument: David Walbert ................. 184
O. Testimony of Linda M eggers......................... 186
P. Maps and Charts
1. Congressional Redistricting Plan MSLSS
(House passed)(Abrams Exhibit 37) . . . . . 190
Ill
Page
2. Congressional Redistricting Plan
AMICUSR (Lewis-GingrichXAmicus
Exhibit 1 ) . . . . . . . . . . . . ........ .................. 192
3. Congressional Redistricting Plan
ABRAMS A (R. 296 and R. 319)............ 194
4. Congressional Redistricting Plan
ABRAMS C (R. 296)................................ 196
5. Congressional Redistricting Plan
ACLU1A (Abrams Exhibit 36) . . . . . . . . 198
6. Congressional Redistricting Plan Plaintiffs’
Remedy 4 (R, 342) ............ . 200
7. Congressional Redistricting Plan Plaintiffs’
Remedy 4X (R. 363) ........... 202
8. Plaintiffs’ Exhibits 47-49 (race
maps from AMICUSR of DeKalb,
Bibb, and Muscogee Counties) .............. 204
9. Chart: Black population in Georgia’s
Eleventh District in F irst Enacted Plan
(DeKalb, Richmond, and Bibb
Counties).................................................... 207
Q. ORDERS noting probable jurisdiction
and consolidating appeals .......................... 208
DATE
7/12/95
8/2/95
8/14/95
8/14/95
U nited States D istrict Court
FOR THE SOUTHEREN DISTRICT OF GEORGIA
A ugusta D ivision
CIVIL ACTION NO. CV 194-008
L ucious Abrams, J r ., et a l ., appellants
v
Davida J ohnson, et a l .
A. RELEVANT DOCKET ENTRIES
NR PROCEEDINGS
243 NOTICE of the State of Georgia’s Calling
of a Special Session of the G eneral
Assembly in Which to Enact a Remedial
Reapportionment Plan by State dfts Zell
Miller, Pierre Howard, Thomas Murphy,
Max Cleland.
251 ORDER setting hearing on redistricting
remedies at Augusta on 8/22/95 at 9 a.m.,
inviting original parties to submit sug
gestions not to exceed 25 pages no later
than 8/15/95, and informing parties that
all pending motions and m atters relating
to costs and atty fees will be considered
(signed by Judge Dudley H. Bowen Jr.),
copies served. [Entry date 08/03/95]
258 P L A IN T IF F S ’ SU B M ISSIO N on
Remedy Per Court Order of August 2,
1995 [edit date 08/14/95]
260 MOTION by A bram s in te rv en o rs for
Reconsideration of [251-1] order dtd 8/2/95
1
2
8/15/95
8/15/95
8/15/95
8/17/95
8/22/95
to allow interveners to submit remedies
for consideration, and to stay Ct.’s order of
8/2/95 to allow the Ga Gen Assembly to
complete congressional redistricting with
brief in support. [Entry date 08/15/95]
262 SUBMISSION of Defendant Thomas B.
Murphy in Response to the Court’s Invita
tion to Submit Suggestions as to Remedy.
267 MOTION by United States to Vacate [251-
1] order of 8/2/95 setting hearing on redis
tricting remedies at Augusta on 8/22/95 to
allow for a legislative remedy with brief in
support. [Entry date 08/16/95]
271 R E SPO N SE by dfts Miller, H ow ard,
Cleland to ct’s 8/2/95 [251-1] order setting
hearing on redistricting remedies. [Entry
date 08/16/95]
274 AMENDED BRIEF of Amici Curiae by
John Lewis and Newt Gingrich.
— Motion hearing held * * *. All pending
motions to intervene DENIED; to appear
amicus curiae GRANTED for the pu r
pose of m aking w ritte n subm issions,
D E N IE D for purpose of m aking oral
arguments or participating in open Court;
* * * [E n try date 09/28/95] [Edit date
09/29/95] Motion to dismiss Abrams inter-
venors is DENIED; Abrams intervenors’
motion to be allowed to submit proposed
redistricting remedies is GRANTED; all
motions to stay or vacate the order call
ing for today’s hearing DENIED; * * *
deadline Oct 15 set for new plan; * * *
3
8/31/95
9/1/95
9/1/95
9/5/95
9/13/95
9/15/95
9/28/95
10/11/95
10/13/95
Abrams intervenors not to participate [in
trial on 2d District]; Abrams intervenors’
proposal for plan of rem edies of 11th
Cong. District to be filed within 10 days;
Counsel to notify the Court in writing if
reapportionm ent plan not passed. * * *
[Entry date 09/28/95] [Edit date 09/29/95]
295 STATUS REPORT by Zell Miller, Pierre
How ard, Max Clelancl as to proposed
Congresional plan [Entry date 09/01/95]
296 PROPOSED REDISTRICTING PLANS
by Abrams Intervenors.
297 NOTICE by Thomas Murphy with respect
to legislative proceedings (status of special
session of Georgia General Assembly)
300 STATUS REPORT by Zell Miller, Pierre
Howard, Max Cleland regarding proposed
Congressional plan passed by Georgia
Senate on 8/31/95 [Entry date 09/07/95]
306 NOTICE by dft Thomas Murphy with
respect to legislative adjournment.
307 STATUS REPORT by Zell Miller, Pierre
Howard, Max Cleland [Entry date 09/18/95]
— Bench trial set for 10:00 10/30/95 before
Chief Judge Edenfield; C ircuit Judge
Edmondson and District Judge Bowren at
Augusta. [Entry date 10/02/95]
319 SUPPLEMENTAL DECLARATION of
Selwyn Carter
321 B R IEF on remedy by plffs [Entry date
10/25/95]
4
10/18/95
10/20/95
10/20/95
[sic]
10/27/95
10/27/95
10/27/95
10/27/95
10/30/95
[sic]
325 ORDER directing parties to subm it a
plan that makes the least chgs, in terms
of line drawing, bringing the 11th Dist.
into compliance w/US Constitution, and a
plan that makes the least chgs, in terms of
line drawing, in GAs present congressional
plan bringing the 11th and 2d Dists. into
compliance w/US Const; the parties are
instructed to confer regarding these plans;
signed by Judge James L. Edmondson);
copies served. [Entry date 10/25/95]
329 ORDER directing parties to submit plans
based on first plan that GA submitted to
Dept, of Justice for preclearance signed
by Judge Dudley H. Bowen Jr.); copies
served. [Entry date 10/25/95]
— Bench trial held as to 2d District; ruling
reserved; rem edy phase to continue on
10/31/95
343 SUBM ISSION of dfts Miller, Cleland,
and Howard in connection with the issue
of remedy.
344 NOTICE of filing minimum change plan
and brief by Abrams Intervenors
345 RESPO N SE/REM ED Y submission by
defendant Thomas Murphy to [329-1] order
and [325-1] order of 10/17 and 10/20/95.
348 BRIEF on remedy by United States
349 BRIEF on remedy by United States
10/[30]- Bench trial held; counsel argum ent and
5
31/95
11/9/95
11/16/95
11/22/95
11/22/95
11/22/95
11/22/95
11/30/95
evidence on remedy phase concluded; rul
ing of the Court reserved; the Court to
notify the parties if additional briefing
permitted. All evidence from 11th District
trial made a part of the record in these
proceedings. * * * [Entry date 11/01/95]
363 ORDER directing plffs’ counsel to submit
proposed plan to be known as plffs’ rem
edy 4X by 11/22/95, comments from each
p a rty not to exceed 12 pgs in leng th
(signed by Judge Dudley H. Bowen Jr.);
copies served. [Entry date 11/10/95]
368 ORDER inviting parties to submit addit
material to indicate chgs in racial composi
tion of DeKalb Co, GA, from the ‘90 census
to present ( signed by Judge Dudley H.
Bowen Jr. for 3- judge ct); copies served.
369 RESPONSE by United States to plffs’
[367-1] response to Ct’s order of 11/9/95,
plffs’ plans remdy 4X and 4X-R
370 RESPONSE/REPLY by Abrams inter-
venors to [367-1] plffs’ response to Ct
order of 11/9/95
371 RESPONSE/SUBMISSION of additional
dem ographic evidence by dfts Miller,
Howard and Cleland to [368-1] order of
11/16/95
372 R E PL Y /R ESPO N SE by amicus John
Lewis and amicus Newt Gingrich to plffs’
suggested [367-1] response/rem edy 4X
and 4XR [Entry date 11/27/95]
374 OBJECTION by S ta te dfts Miller,
6
12/1/95
12/1/95
12/1/95
12/6/95
12/7/95
12/8/95
12/8/95
Howard, Cleland to [369-1] response by US
Dept of Justice to plffs’ plans 4X and 4XR.
375 REPLY by plffs to [369-1] response by
United States, [370-1] response by Karen
Watson, William Gary Chambers Sr., G.
L. Avery, Lucious A bram s Jr., [372-1]
response by Newt Gingrich, John Lewis
to court order of November 9
376 RESPONSE memorandum by Abrams
intervenors to [371-1] submission of addi
tional dem ographic evidence by Max
Cleland, Pierre Howard, Zell Miller
377 ORDER that Georgia’s Second Congres
sional D istrict is unconstitutional in its
current composition; dfts are barred from
using it in future congressional elections;
(signed by Judge Jam es L. Edmondson
for three-judge court); copies served.
378 RESPONSE by United States to [374-1]
objection by Cleland, Howard and Miller,
and response to [375-1] response by plffs
to US subm ission of plan on 11/22/95
[Entry date 12/07/95]
379 OBJECTION by defendants Miller,
Howard, and Cleland to [372-1] reply/re-
sponse by Amici Curiae Gingrich and Lewis
to plffs’ suggested remedy 4X and 4XR
380 OBJECTION by plffs to submission of
affid of Donald Hill in [372-1] response by
Newt Gingrich and John Lewis
381 RESPONSE by United States to [371-1]
submission of additional demographic
information by Cleland, Howard, Miller
7
12/13/95
12/13/95
1/11/96
1/11/96
1/12/96
384 R E S P O N S E of am icus Lew is and
Gingrich to [379-1] objection by state dfts
C lei and, Howard, Miller to reply of amici
to plffs’ 4X and 4XR suggested remedy,
and response to [380-1] plffs’ objection to
amici’s submission of affid of Donald Hill
385 ORDER redraw ing Ga’s congressional
districting plan; directing tha t elections
for members of the House of Reps of the
Congress of the US from the St of GA be
conducted in accordance with the plan
appended to this Order as Appendices
“A” and “B”; motions terminated; signed
by three-judge court Edenfield, Bowen
and Edmondson; copies served. [E ntry
date 12/14/95]
397 N O TIC E OF A P PE A L by Lucious
Abrams Jr., G. L. Avery, William Gary
Chambers Sr., Karen Watson to Supreme
Court of the US from Order dtd 12/13/95,
oral ruling on 8/22/95, and Order of 1/8/96;
copies served. [385-1] order, [395-1] order
[Edit date 01/12/96]
398 MOTION by Lucious Abrams Jr., G. L.
Avery, William Gary Chambers Sr., Karen
Watson for a Stay pending appeal of Order
of Ct entered 12/13/95, permitting ‘96 pri
m ary and general elections to be held
under a redistricting plan that does not
violate Sec. 2 and 5 of Voting Rights, with
brief in support. [Edit date 01/16/96]
401 NOTICE OF APPEAL by dft-intervenor
United States to the Supreme Court of the
8
US from 12/13/95 order of USDC; copies
served. [385-1] order [Entry date 01/16/96]
1/26/96 407 ORDER denying Abrams In tervenors’
[398-1] motion for a Stay pending appeal
of Order of Ct entered 12/13/95 (signed
by Judge J. L. Edm ondson, Judge B.
Avant Edenfield and Judge Dudley H.
Bowen J r . ); copies served.
9
B.
TESTIMONY OF JOSEPH KATZ
[Excerpt from Trial Testimony, 5 T. Tr. p. 63,
line 5 to p. 86 line 1,]
Could you tell us, in a more conversational way, what
you have found. Essentially, to this point, you’ve been
asked to undermine Dr. Lichtman’s work, I think. So tell
us why it’s not worthy of your credit.
MR. WALBERT: I was just moving to the next point,
Your Honor.
HONORABLE JUDGE BOWEN: Tell me, in a more
conversational way, about his work, what you find trou
bling about it?
A (Dr. Katz) The thing I find troubling about it is
attempting to estimate election by election the actual
voting percentages of the white voters for the white
candidate and the black voters for the black candidate.
The data that he has available, that any of us have
availab le, is m erely th e re g is tra tio n d a ta for the
precinct, and the vote data for the precinct. And if we
have precincts that are racially mixed in that there’s a
high percentage of white and black voters, with only the
vote data and the registration data, it really isn’t possi
ble to dig out what the percentages of racial voting pat
terns are. And this is magnified when we had very few
precincts to work with.
HONORABLE JUDGE BOWEN: Could it be said
that any time you have a precinct in which there are two
races voting, that there is inevitably some extrapolation
in saying who voted and who did not?
A (Dr. Katz) Yes. And to come up with an estimate by
necessity, you have to make some additional assump
tions about the way the different races voted.
10
HONORABLE JUDGE BOWEN: You asked, or your
[sic] mentioned, some impossible results. Are those
results that are impossible because of mathematical cal
culation or because of the methodology used?
A (Dr. Katz) In these cases, i t ’s because of the
methodology used. The quadratic relationship that I
have demonstrated here, if you try to start fitting linear
equations to election data where your precincts fall, say,
under 40% black registered voters, then what you’re
getting in effect are observations that fall somewhat
along that line where turnout increases.
And if you then fit a line to that, that will simply shoot
the turnout further up, and it never has a chance to
adjust back down.
HONORABLE JU D G E BOWEN: Go ahead, Mr.
Walbert.
MR. WALBERT: If Your Honor doesn’t mind, if I
could make an inquiry of the Court, I just found out that
one of the copies I was working off of this report when
the p rin te r made it apparently left all the exhibits
out behind Tab Five, and if I might just to make sure
that —
HONORABLE JUDGE BOWEN: have - - -
MR. WALBERT: Do you have something?
HONORABLE JUDGE BOWEN: I have charts and
graphs on mine.
MR. W ALBERT: All rig h t. T h a t’s good. I ju s t
checked the original one that is marked as an exhibit in
evidence as well.
HONORABLE JUDGE BOWEN: Well, Dr. Katz, I
know you and Mr. Walbert have talked about this a num
ber of times, and that there are necessary terms of art,
11
but don’t hesitate to speak to us in a conversational way.
Q (Mr. Walbert) Now, Dr. Katz, I want to turn your
attention to the ultimate question here in this Court
about the likelihood of this kind of evidence, the likeli
hood of determining whether a black candidate would or
would not win in a district of a certain percentage of
black voters.
And let me ask you, before we turn to your calcula
tions and your charts, does Dr. Lichtman’s methodology
allow one to make those kinds of predictions or not?
A Not as methodology to the extent that he merely
estimates the racial polarization of voting estimates.
Q All right. Now have you used statistical techniques
and standard analytical techniques, Dr. Katz, to try to
make determinations of the likelihood of a black candi
date prevailing in a black-on-white election in congres
sional districts of various percentages, using the data
that Dr. Lichtman relied upon?
A Yes. That data and some additional data.
Q Okay. Now, are the final determination, before we
go back to them in detail, are those the charts that are
contained in your report, most of them anyhow, behind
Tab Five?
A Yes, they are.
Q Now, I’m going to pick out from here, if I could,
what I believe to be the first one of those charts behind
Tab Five, and put it up here.
Now using the data that Dr. Lichtman has, and he
relied upon, tell the Court what is depicted here by the
probability of the black candidate winning an election,
what this line and data plot would allow one to use this
chart for.
12
A Okay. What I’ve tried to represent here is a fre
quency distribution of the percentage of precincts in all
of the elections in the database, in which the black can
didate or black candidates had the majority of the vote.
Q Let me ju s t get you, what is, if you wanted to
determine, Dr. Katz, the likelihood of a black candidate
winning in a black and white race, can this chart, based
on the database here, be used for that purpose?
A Yes.
Q How would you do that?
A You would look at the racial composition of the
election. Say there’s 25 to 30% black registered voters,
and the estimate, or the probability that the black candi
date would win is the number approximately 20% in this
box here.
HONORABLE JUDGE BOWEN: Do you make any
assumption about the qualifications of the candidates
being equal?
A No. W hat I’m trying to do is just get an overall
probability without limiting it to —
HONORABLE JUDGE BOWEN: This is your, this is
your pro jection , not L ich tm an’s, you’re using the
Lichtman information. Is that right?
A I’m using the database from Dr. Lichtman, but I’m
not using it in the way he used it.
HONORABLE JUDGE BOWEN: All right. This is
your graph.
A Yes, sir, that’s correct.
HONORABLE JUDGE BOWEN: Okay.
Q (Mr. Walbert) Now, let me, I think, flip a couple
13
more charts back in your exhibit Tab Five there , I
believe this would be the one that is depicted there, Dr.
Katz.
In this graph, is it the same kind of presentation of
the graphical analysis that was on the previous one?
A Yes, it is. What’s changed in the database of elec
tions that I’ve used to conduct the analysis.
Q How has the database been changed in preparing
this particular one, Dr. Katz?
A In this graph, I enhanced the Lichtman database to
include the r e s t of th e p rec incts from th e 1990
Governor’s primary.
Q Let me stop you there, if I might. When you say
you enhanced it to include the rest of the precincts for
the 1990 Governor’s primary, what do you mean?
A Dr. Lichtman’s database did not contain all of the
precincts from the 1990 Governor’s election, primary
election.
Q Was there any apparent methodology to his select
ing ones here and there that you could determine at all?
A Not that I could determine.
Q Was Dr. L ich tm an’s da tabase , pu t aside the
Governor’s election, but the rest of the election, was
there any geographical limitation to it, or was it from
various places all over the State of Georgia?
A Mostly from all over Georgia, although in predomi
nantly white counties, I don’t think he had any election,
probably because there weren’t any black/white elec
tions.
Q All righ t. Now, his database, in fact, was ju s t
14
black/white elections; wasn’t it?
A That’s correct.
Q And was it all democratic primary elections except
for a couple of judicial ones?
A There were judicial elections, there was also a few
general elections. But I didn’t use those.
Q All right, sir. He testified in [h]is deposition that
primaries were the ones that were most important to
him, Dr. Katz?
A That’s my understanding from reading of his depo
sition.
Q Okay. Is that why you used the primaries when you
ju st said you did?
A Yes.
Q Okay. Now, you have also enhanced Dr. Lichtman’s
database you say, you just said, by using the rest of the
Governor’s race that he is a part of. Why is the 1992
Labor Commissioner and the 1992 Second and Eleventh
Congressional Districts primary elections in that data
base as well?
A Those are additional black/white elections in which
I can enhance the overall number of precincts to get
more precise estimates of the probabilities of the black
candidate prevailing.
Q Okay. Now, was the selection of those black/white
democratic primary elections consistent with the rules
that Dr. Lichtman has purported to you in the develop
ment of his database, Dr. Katz?
A Yes.
Q Were you making any independent political judg-
15
ments about the propriety of elections or just trying to
track his method, his selection?
A I was trying to track Dr. Lichtman’s selection.
Q Now —
HONORABLE JUDGE EDMONDSON: This is a lit
tle bit unusual, I guess, because we’re hearing a witness
whose purpose is to undercut the testimony of another
witness who we have not yet heard.
Explain to me why it is, what is the point that you and
the government of the United States are disagreeing
about that is pertinent to the outcome of this case.
I understand that there is a great deal of disagree
m ent betw een the S tate of Georgia and the United
States Government about elections in Georgia in gen
eral, and other kinds of elections, but, for our case, what
is the point of controversy that I should be focusing on
here?
MR. WALBERT: I think this, Your Honor, as far as 1
understand Dr. Lichtman’s testimony, when you start
going down this kind of a chart here, it is my under
standing that his opinions, as he has ventured before, is
that you, basically, blacks can’t win under some level—
and Fm not sure that he has ever quantified that specifi
cally, but I would expect that he would in this case on
his testimony—and what we are showing here is that
there is a whole lot different probability and pattern, if
you will, Judge, than I expect Dr. Lichtman to testify to.
I think he has basically got his concept about, if Fve
read him correctly in other case, is that there is kind of a
step function here. You’re in a safe d istric t a t some
point, and then it rapidly drops off into oblivion and it
stays down there. And I think that’s the critical question.
16
HONORABLE JUDGE EDMONDSON: I did under
stand from reading summaries of these reports that that
was where, that there was a difference of view on that
point.
What I do not understand is how that m atters to this
case, since whatever the percentage of voting-age black
vo ters are in the E leven th D istric t or the Second
District is, I gather, not now disputed, and what is the
significance of this pointing out that Georgia thinks one
thing and the United States Government thinks some
thing else when what is being done here is defending a
district that already exists?
MR. WALBERT: Well, to some extent, Your Honor, I
guess it’s in response to Your Honor’s instructions or
suggestions, and I think it is an appropriate one, that we
are the client of the law in this case in the highest and
best sense.
And I think it would be an inappropriate thing to
come into this Court and to defend this district to say
and to take the position that it is absolutely necessary
for [a] black person to have any chance of election that
this district being the percentage that it is.
And I th in k th e go v ern m en t’s ev idence of Dr.
Lichtman tends to be in that direction. I don’t think, the
State has not taken that position. And I think that what
we ought to try to do here is just—we have no idea how
this is going to turn out, quite frankly, when Dr. Katz
started doing it.
We just said, my request to Dr. Katz was: Is there a
way that you can statistically, with statistical quantita
tive methodology, come up with this. With an ideal of a
real, what’s the likelihood of winning. Because that’s the
bottom line.
17
HONORABLE JU DGE EDMONDSON: Am I to
understand that the chief reason that you’re litigating
this point in this case is to prevent th is Court from
including in its decision in this case some statement that
a State black district must have “X” percentage of vot
ers because you don’t want to have that precedent hang
ing over you in other litigation?
MR. WALBERT: I think tha t’s one, that has partly
stated one of the two reasons.
HONORABLE JUDGE EDMONDSON: What is the
other one?
MR. WALBERT: The other reason really is to show
to the Court what the truth is, if you will, the best that
we can, in terms of what percentages are appropriate to
the likelihood of black candidates winning in a district of
different percentages. To me, that is a very germane
point in this case, to our way of thinking.
And I think tha t the State—I guess that there’s—
Judge if I can focus in real closely, I do think that the
intervenors and the government take the position of
both se ts of in te rv en o rs th a t the S ta te was really
absolutely compelled as a matter of law under Section 2
to enact this district. We don’t take that position. I t’s a
fundamentally different position.
We take the position that the State having enacted it,
giving this type of chart here and these type of results,
it was reasonable w hat the S ta te did in its overall
action, and that the Constitution should not strike down
what the State did.
But we fundamentally disagree, Your Honor, on the
fact th a t th e S ta te was com pelled by e ith e r the
Fourteenth Amendment or Section 2 to do this.
18
HONORABLE JUDGE EDMONDSON: This ju st
goes to justifying the district if it is found to be bizarre
under Shaw; is that right?
MR. WALBERT: Yes, I think—yes, I think tha t’s a
fair statement, yes sir.
HONORABLE JU DGE EDMONDSON: I under
stand this much better now.
MR. WALBERT: I appreciate it. Thank you.
Q (Mr. Walbert) I think we’ve probably covered that
exhibit as much as need be, Dr. Katz. Let me ask you, if
I might, one last point which goes to the first reason
here, and that is this:
Did Dr. Lichtman also include in his database, in his
analysis, some of the judicial elections in Georgia, such
as some precincts from the R obert Benham race in
1984?
A Yes, he did.
Q Have you made an assessment, Dr. Katz, of all of
the database on judicial elections in his database, with
all of the precincts, with all of the state-wide precincts
for Judge, now Justice, Benham, and Justice Sears-
Collins, in all of the four state-wide elections?
A All four state-wide, yes, I have.
Q All right.
A And one of these exhibits shows the probability of
the black candidate winning the election if we throw in
these additional elections.
Q Now, you used the judicial elections on the local
level that Dr. Lichtman had as well; didn’t you?
A Yes.
19
Q Did you m ake a finding or d e te rm in a tio n of
whether the judicial elections are, in fact, sufficiently
similar in their characteristics statistically, to be a reli
able indicator in this kind of race, or are they materially
different?
A I found that statistically, the judicial elections are
materially different.
Q Would it, from a statistical point of view, be inap
propriate in your judgm ent to rely on the one group
going back and forth from congressional to judicial or
judicial to non-judicial elections?
A Inappropriate in the sense that they seem to be
measuring different kinds of elections.
MR. WALBERT: We have no further questions of this
witness, Your Honor.
HONORABLE JUDGE BOWEN: Dr. Katz, just one
other thing. This graph that you have here, do you pre
sent this as a projection of what would, or what might
happen, or simply a report of what has happened in the
past?
A (Dr. Katz) I presented it as what might happen,
since this was the only empirical evidence that I can col
lect to try and predict the voting pattern in the future.
HONORABLE JUDGE BOWEN: There is no rela
tionship, as far as you know, to the qualifications of the
candidate? It is strictly race-based?
A (Dr. Katz) Yes.
HONORABLE JU D G E BOW EN: All rig h t. Ms.
Murphy, you may have a question or two.
MS. MURPHY: I have a few minutes of examination
20
for Dr. Katz. Fm wondering if the Court would like to
take it’s morning break now, before I start?
HONORABLE JUDGE EDMONDSON: How much
time is it going to take?
MS. MURPHY: I expect it will take about tw enty
minutes.
HONORABLE JUDGE EDMONDSON: W hat do
you think, Judge Bowen? We would like to take the
morning recess about 11:00, which gives you about sev
enteen minutes. Let’s see if you can do your twenty in
seventeen.
MS. MURPHY: Well, I hope you will give me that
extra minute or two if I need it.
HONORABLE JUDGE EDMONDSON: Well, keep
in mind that speed is the essence of war.
MS. MURPHY: I hate to think we’re actually at war
here.
HONORABLE JUDGE BOWEN: Go.
CROSS EXAM IN ATIO N B Y
MS. MURPHY:
Q Dr. Katz, I’m Donna Murphy, good morning.
A Good morning.
Q And, as you know, I represent the United States.
You’ve read Dr. Lichtman’s entire report in this case; is
that correct?
A Yes.
Q And the analysis that you were describing that is
21
set forth on pages 1 to 18 of your report, that analysis is
based only on the data that underlies the appendices of
Dr. Lichtman’s report; is that correct?
A That’s correct.
Q So, Dr. Lichtman did, in fact, analyze the elections
you were talking about a few minutes ago, the 1990
gubernatorial primary, the 1990, the 1992 labor commis
sioner primary, and the 1992 congressional election in
Districts Eleven and Two, in the table in the body of this
report.
A Yes, but they were just partial, it wasn’t the com
plete election.
Q B ut you analyzed them for th e E lev en th
Congressional and the Second Congressional Districts?
A Yes.
Q And so, you[r] analysis of the statistical methodol
ogy and whatever data issues that you have would only
go to the appendices of Dr. Lichtman’s report. Is that
correct?
A That’s correct.
Q With regard to those appendices, in Dr. Lichtman’s
report, he doesn’t rely on any one election in those
appendices to draw a conclusion; does he?
A I don’t know what he relies on.
Q Well, you’ve read his report. Does he purport to
rely on any particular election percentages?
A Not only does he not rely on any particular one, he
never provides an overall assessm ent of how all the
results for all the elections fit together in coming up
with an estimate, and overall estimate.
22
Q So he basically uses those as sort of a background
to show some, over 300 elections I believe, what his
assessment of racial voting patterns is in Georgia over a
wide number of elections. Is that correct?
A That’s my understanding as to his intentions, yes.
Q And, have the pages tha t you have attached to
Tabs Three and Four of your report, from Exhibits
“168” and “169”, you didn’t pull those pages out, you
pulled those pages out to illustrate the problems that
you found; is that correct?
A That’s correct.
Q So, these are going to be the ones that show what
the problems might be, not necessarily representative of
the typical page in your report of Dr. Lichtman’s analysis.
A As part of my report, I also give overall summary
statistics to the extent to which these problems existed
in this set of elections.
Q But my question was with regard to these particu
lar pages that you pulled out, those are designed to illus
trate the problem; is that correct?
A Yes.
Q Now with regard to your analysis on pages one
through, I believe it is 18 of your report, it basically
boils down to a critique of the use of double ecological
regression analysis to analyze racial voting patterns;
isn’t that correct?
A I’m not sure I understand the reference that you
gave me. Page 18?
Q Page one through 18, I’ve been talking about this
whole first section —
23
A Oh, one through 18.
Q of your report.
A Yes.
Q Are you familiar with the Supreme Court opinion
of Thornburg v. Cringles!
A I read that opinion, it’s been a while, yes.
Q And you’re aware tha t in tha t case, the Court
relies on, in fact, double ecological regression analysis
estimates of voting behavior in drawing its conclusions
about racially polarized voting.
A I don’t recall whether they relied on double ecolog
ical regression estimates, but I think they did rely on
some ecological regression estimate.
Q And those ecological regression estim ates were
performed by Dr. Bernard Grofman; is that right?
A I’m not sure.
Q You don’t have any reason to dispute that, though?
A No.
Q Are you aware of any differences betw een the
m ethodology th a t Dr. L ichtm an em ploys and th e
methodology that was employed by the plaintiffs’ expert
in the Thornburg v. Gingles case?
A I’m not familiar with the methodology employed in
the Thornburg v. Gingles case, so I can’t, you know, tell
you if there is any differences.
Q So if I —
A But I also don’t know the extent to which that data
had more precincts more amenable to double ecological
regression.
24
Q But if I would represent to you that Dr. Lichtman
employed the same analysis as Dr. Grofman did in the
Thornburg v. Gingles case, you wouldn’t have any rea
son to dispute me on that, would you?
MR. WALBERT: Objection as to the form of tha t
question.
MR. PARKS: The plaintiffs join in that. I don’t want
to hear any objections, but it’s —
MR. WALBERT: But he has ju s t testified th a t he
didn’t know what happened, and for her to ju st argue
the case and say, “you can’t then tell me that — ”
HONORABLE JUDGE BOWEN: “Well, that might
be argumentative, and we’ll save that.
Q (Ms. Murphy) Are you aware, Dr. Katz, that in the
case of Garza v. County o f Los Angeles, the type of cri
tique that you’re presenting here was presented exten
sively by a number of different statistical experts con
cerning analysis performed by both Dr. Lichtman and
Dr. Bernard Grofman?
A Yes, but your, you have to keep in mind that the
election data makes a big difference as to the appropri
ateness of methodology —
Q I am saying —
A The methodology, let me finish, please.
Q I’m sorry.
A A methodology by itself isn’t usable for every con
ceivable data set. Just to say that somebody used eco
logical regression in one circumstance and it turned out
to be a reasonable methodology, doesn’t make it reason
able in every circumstance.
25
Q So you’re saying that the Court in the Garza case
rejected the kind of criticism that you’re offering that
that wouldn’t necessarily lead to the same conclusion
concerning your critique of Dr. Lichtman’s analysis.
A Not necessarily. The data is totally different.
Q And are you aware that the Garza Court did reject
those critiques?
A I’m not—I haven’t read the Garza opinion, and I
don’t know what the Court did.
Q Now, after, on page, I believe, 20 of your report,
you get into a som ewhat different area, where you
employ your own, if I can call it, hom ogen[e]ous
precinct analysis to try and estimate racial voting pat
terns; is that correct?
A Average racial voting patterns.
Q Average racial voting patterns.
A Yes.
Q And if I understood your earlier testimony, you put
three different sets of estimates for three different data
bases; is that correct?
A Is this on page 20 and 21?
Q Pages 20 through 21.
A Yes, I do.
Q And if I understood your earlier testimony, I take
it th a t the th ird e s tim ate th a t you rep o rt, which
includes judicial elections, you would consider to not be
very relevant or germane to this case. Is that correct?
A Well, in terms of relevance or germane to this case,
I don’t have an opinion on that. I t’s there to show that
26
when you throw in the additional judicial elections, that
there is a substantial increase in the average percent
vote of whites voting for the black candidate.
Q I apologize. I th ink my question was poorly
phrased. In fact, you found a different pattern in the
judicial, the state-wide judicial election that you ana
lyzed in terms of voting patterns; is that correct?
A Different from what?
Q Different from the pattern you found in, say, the
number two, all the data you analyzed in the number
two estimate.
A Yes.
Q So which would you consider if you were trying to
estimate voting behavior in a congressional election,
congressional primary election. Which of these three
sets of estimates that you report would you consider the
most, the best estimate?
A Following Dr. Lichtman’s political assumptions? Or
do you want me to make my own assumptions, what I
think would be the best thing to use.
Q Well, what assumptions would you make? I mean,
your not a political scientist, are you?
A No.
Q Okay, so what assumptions would you make? Let’s
take a step backwards.
A Well, from a statistical perspective, I would want
Congressional elections to estimate the degree of racial
polarized voting in future Congressional elections, and
other elections that are related to that.
Q So I’m asking you of these three sets of estimates
27
you report here, which would you consider the best and
the most reliable. I realize they’re not perfect, I under
stand your testimony.
A Okay. I would go with number two.
Q And based on the set of estimates in number two,
would you say that there are substantially or signifi
cantly different voting patterns among black and white
voters according to these estimates?
A Yes, in that the competence interval surrounding
these numbers are about, plus or minus, 1%.
Q And do you have an opinion on w hether or not
these numbers would be indicative of racially polarized
voting? Do you have an opinion?
I don’t know if you’re, if you feel comfortable giving
an opinion since your [s?'c] a statistical person, not a
political scientist or a political analyst.
A To the degree of whether it’s legally significant,
o r - - -
Q No, just, if you understand the term racially polar
ized voting do you have an opinion on these numbers?
A Well, my opinion is that there is a 95% competence
interval to the average percent white voting for black
candidates that goes from 27% to 29%. So, to the degree
that whites tend to vote for white candidates in the
range of the approximately 71 to 73%, that’s, there is a
tendency for white voters to vote for white candidates.
Q Maybe I should ask the question a little differently.
Rather than getting into a term that obviously has some
legal implications, these numbers indicate a fair degree
of cohesion among white voters and cohesion among
black voters; is that correct? Can you agree with that?
28
A By cohesion, I think that’s a legal term, too.
Q Oh, is it? Voting similarities? Can we agree on a
term?
A W hites tend to vote for w hite candidates and
blacks tend to vote for black candidates according to the
results of number two.
Q Thank you. I have a question—I’m not sure if you
still have these charts in front of you. This exhibit which
you’re still looking at here, Plaintiffs’ “174”, and you
indicated that this is a chart indicating the probability of
a black candidate getting elected where the percentage
of black voters fell at certain intervals; is that correct?
A Yes.
Q And this particular chart we’re looking at is the,
includes the 1990 gubernatorial and 1992 labor commis
sioner, and the 1992 Second and Eleventh Congressional
D istrict races, as well as the many smaller elections
included in Dr. Lichtman’s data; is that right?
A Yes.
Q According to your estimates as based on this chart,
a t approxim ately—will you agree w ith me th a t the
probability of a black candidate winning doesn’t go
above 50% until the percentage of black registered vot
ers reaches approximately 50%?
A Yes. In the range from 45 to 50, that percentage is
a 46% probability, and from 50 to 55%, that’s a 62% prob
ability. It’s about 50%.
Q And, I’m now showing you Defendants’ Exhibit
“172”, which is a similar chart that you did based only on
what you termed “the Lichtman database,” is that cor
rect?
29
A That’s correct.
Q And if I asked the same question, at what point
does the probability of a black candidate winning go
above 50%, what would your response be? Is it approxi
mately 50%?
A. Let me check my numbers to be sure. Yes, it’s
approximately 50%. It’s 44% in the range from 45 to 50%
black registered voters, and the probability is 55% in
the range from 50 to 55%.
30
C. [CORRECTED] REPORT OF
DR. JOSEPH L. KATZ
[State’s Exh. 170, at 14 (Part II)-22]
HI. A N A L Y SIS OF THE R E L IA B IL IT Y OF THE
E S T IM A T IO N OF R A C IA L VO TING P E R
CENTAGES D E TE R M IN ED B Y DR. LICHT-
MAN.
I have analyzed Dr. Lichtman’s primary election data
base that was provided on computer diskettes by the
D epartm ent of Justice and Dr. Lichtman. The data
included on these diskettes encompass the great major
ity of those elections for which racially polarized voting
estimates are given in Appendix 1 and Appendix 2 of Dr.
Lichtman’s report. I find that the potential difficulties
with double ecological regression racial voting esti
mates, enumerated in the previous section, are present,
to varying degrees, in the analysis of these elections.
Approximately 50% of the elections in the Lichtman
database have nine or fewer precincts. Approximately
38% of the elections have one or more impossible values
for the four turnout estimates. Approximately 32% of
the elections have registration data taken from a year
that is different from the year of the election.
The rac ia l po lariza tion vo ting e s tim a te s in Dr.
Lichtman’s report are single numbers or “point esti
mates” derived as ratios from four estimates of white
and black voter turnout for white and black candidates
in an election, taken from two separate “least squares”
lines. In the context of statistical regression methodol-
ogy, each of the four voter turnout estimates is subject
to some degree of variation from the “tru e ” vo ter
31
tu rn o u t percen tages th a t are sough t.4 I have con
structed 95% confidence intervals for each of the four
voter turnout estimates to assess this variation.5 These
confidence intervals, in turn, assist in evaluating the reli
ability of the racial vote estimates derived from double
ecological regression. My detailed analyses of these
elections, individually, are contained in State Exhibit 166.
(Illustrative excerpts from this exhibit that pertain to
several of the Lichtman elections follow Tab 3)
A brief summary of the results of my confidence inter
val determinations are the following:
(1) In fully half of the elections in the Lichtm an
database, the length of the 95% confident interval
for the white turnout for white candidates was
26% or greater.
(2) In half of the elections, the length of the confidence
intervals for white turnout for black candidates
was 18% or greater.
4H ypo the tica lly , th is p a r tic u la r an a ly s is w ould n o t ap p ly in a case
w h e r e e v e r y p r e c in c t in a n e le c t io n w a s 100% h o m o g e n e o u s ,
w h ite o r b lack , b ec au se in th a t in s tan ce double ecological re g re ss io n
p ro d u c e s e x a c t r a c ia l v o te p e r c e n ta g e s . In th e L ic h tm a n d a ta ,
w h ich is co m p rised e n tire ly of p re c in c ts th a t a r e a t le a s t to som e
e x te n t m ixed , th e re is n e c e ssa rily som e d e g re e o f s ta t is tic a l v a r i
ab ility in v o te r tu rn o u t e s tim a te s .
5I h a v e u sed 95 p e rc e n t confidence in te rv a ls , w h ich m ean s th a t ,
b a se d on th e e lec tio n d a ta ana lyzed , th e re is a .95 p ro b a b ility th a t
th e “t r u e ” v o te r tu r n o u t p e rc e n ta g e fo r a n e le c tio n is co n ta in e d
so m ew h e re in th e in te rv a l. T he 95% level is ty p ica lly u sed in s ta t is
tica l ana ly sis .
T h e s ta t is t ic a l r e g re s s io n m eth o d o lo g y fo r c o n s tru c tin g confi
dence in te rv a ls is n o t a p e rfe c t a n sw e r to th e v a r ia b ili ty issu e w ith
ecological re g re s s io n e s tim a te s , b u t i t is th e m o st a p p ro p ria te m ea
s u re ava ilab le .
32
(3) In half of the elections, the length of the confidence
interval for black turnout for black candidates was
23% or greater.
(4) In half of the elections, the length of the confidence
interval for black turnout for white candidates was
34% or greater.
The sensitivity of the racial vote percentages to the
lengths of these 95% confidence intervals is demon
strated by the following example. Suppose that, in a cer
tain election, double ecological regression predicts a 15%
turnout of whites for the white candidate(s) and a 5%
turnout of whites for the black candidate(s). Therefore,
the estimated percentage of white voters voting for the
white candidate(s) is 15/(15 + 5) or 75%, and the esti
mated percentage of white voters voting for the black
candidate(s) is 25%.
Using the above-stated median percentage for our
example indicates that there is a 95% chance that the
“tru e ” percen tage of w hite reg is te red vo ters who
turned out for the white candidate is between 2% and
28%. Similarly, there is a 95% chance that the “tru e”
white turnout for the black candidate(s) is between -4%
and 14%, even though a negative result is a practical
impossibility.
Different selections of voter turnout from the two
95% confidence intervals can generate significantly dif
ferent white vote percentages. For example, if 6% of the
white registered voters turn out for the white candi
date^) and 14% for the black candidate(s), then 20% of
the white voters still turn out to vote, as in our initial
estimate. However, only 30% of the white voters are
then predicted to have voted for the white candidate(s),
compared to the original 75%.
I also produced 95% confidence intervals, under the
33
statistical regression methodology, for racial vote esti
mates based upon single ecological regression for a sub
set of elections from the Lichtman database.6 For the
most part, these confidence intervals are fairly wide. My
detailed analyses of these elections, individually, are
contained in S ta te E xhib it 165. (Some illu stra tiv e
excerpts from this exhibit that pertain to several of the
Lichtman elections follow Tab 4)
As part of his results, Dr. Lichtman reports squared
correlation coefficients, or R2 values. Although Dr.
Lichtman estim ates racial voting behavior based on
double ecological regressions, the R2 reported by Dr.
Lichtman pertains to values he has calculated for single
ecological regression. More significantly, R2 values do
not m easure the accuracy or degree of confidence of
racial voting estim ates derived from single or double
ecological reg ress io n .7 Thus, a high R2 value says
nothing about the statistical confidence level applicable
to those voting estimates.
I conclude that Dr. Lichtman’s small election by small
election analyses do not provide a meaningful and reli
able description of racial voting behavior in Georgia.
IV. A N A L Y SIS OF RAC IAL VOTING P E R C E N T
AGES OVER M A N Y GEORGIA ELECTIONS.
While Dr. Lichtman’s analyses do not lend themselves
to a meaningful description of racial voting behavior in
6T h is s u b s e t w a s t h e p o r t i o n o f h is d a t a p r o v id e d b y t h e
D e p a r tm e n t o f J u s tic e in an e a r lie r d isk e tte .
7F o r e x m p le , i f a ll th e p re c in c ts in th e e le c t io n w e r e 100%
racia lly hom ogenous, s in g le and double ecological re g re s s io n w ould
p roduce ex a c t ra c ia l v o tin g p e rc e n ta g e s re g a rd le s s o f th e slope of
th e le a s t s q u a re s line. B u t th e R 2 v a lu e could b e a n y w h e re from 0
to 1.0
34
Georgia generally, that is not to say that this data is
utterly irrelevant to that issue. A better methodology
using this data would be to measure overall racial voting
behavior by estimating the average racial voting per
centages over a representative set of Georgia elections.
My method differs from that of Dr. Lichtman and Dr.
Weber in that I do not estimate racial voting patterns
in individual elections, but seek to examine average vot
ing patterns over a number of elections.8
Moreover, because of the methodological problems
inherent in double ecological regression analysis, and
particularly because of the incorrect assumption that
voter turnout is a linear function, a methodologically
sound way to determine estimates of racial voting pat
terns is to look at the homogeneous precincts—i.e.,
those 100 percent one race or the other, or nearly so.9
To determine overall voting behavior, the results for all
of the homogenous precincts for the various elections
8W hile th e se m ethodo log ies can e lim ina te som e o f th e s ta tis tic a l
p ro b le m s of Dr. L ic h tm a n ’s an a ly s is , th e y s ti ll do r e q u ir e c e r ta in
a s su m p tio n s a b o u t v o tin g behav io r. F o r ex a m p le , a h om ogeneous
p re c in c t an a ly s is r e g u ir e s th e a s su m p tio n th a t w h ite v o te r s in a
hom ogenous w h ite p re c in c t v o te th e sam e as w h ite v o te rs in a non-
h o m o g en o u s p re c in c t, an d th e sam e th in g fo r b la c k v o te r s . T h a t
m ay o r m ay n o t b e t ru e , and i t is s ta tis tic a lly im possib le to d e te r
m ine w h e th e r o r n o t i t is tru e .
9In th is ana ly sis , I h av e u sed only th o se p re c in c ts th a t a re 95%-
100% one race o r th e o ther. Dr. L ich tm an does do a s im ila r k ind of
h om ogenous p re c in c t ana ly sis , b u t h is d e te rm in a tio n s c o n sid e r as
“h o m o g e n e o u s” a n y p re c in c t th a t is b e tw e e n 80% an d 100% one
r a c e o r t h e o th e r . C a ll in g p re c in c ts “h o m o g e n e o u s ” w h e n th e y
h a v e such a la rg e d e g re e o f h e te ro g e n e ity su b s ta n tia lly a ffec ts th e
re lia b ility and accu racy o f th e e s tim a te s d e r iv e d .
35
are averaged together, weighted by the number of vot
ers in the precincts.10
I have also made these determinations of racial voting
patterns using homogenous precinct analysis for several
different data sets. F irst is the Lichtman data set, again
utilizing all of those elections of his that were provided to
me by Dr. Lichtman and the Department of Justice. I
have also made the same determinations for an election
data set consisting of those in Dr. Lichtman’s data set
plus other major recent elections involving black-white
candidates in Georgia—the Governor’s primary election
in 1990 (which involved Andrew Young; only a portion of
the precincts from this election were included in Dr.
Lichtman’s data); the 1992 Labor Commissioner primary
(which included black candidate A1 Scott); and the sec
ond and eleventh congressional primary races from 1992.
Finally, I have made the same calculations for the
most comprehensive available database of Georgia elec
tions with black-white opponents. This includes the
Lichtman data, all of the four large races mentioned
above, and all four state-wide black-white judicial races
(again, portions of one of these races were already
included in Dr. Lichtman’s data).
The results are as follows:
i°“W eig h tin g ” g iv es a p ro p o rtio n a lly g r e a te r w e ig h t to a p re c in c t
w ith m o re v o te rs th a n one w ith less. T his avo id s a n o th e r p ro b lem
in h e re n t in th e p re s e n ta tio n b y Dr. L ich tm an w h ich m ak es no d is
t in c t io n b e tw e e n a n e le c tio n th a t m a y h a v e o c c u r re d in a sm all
co u n ty o r d is tr ic t w ith a few v o te rs an d one th a t m ay h av e o ccu rred
from a n elec tion in a c o u n ty o r a d is tr ic t w ith fa r m o re v o te rs .
LICH TM AN ELECTION DATA (No. 1)
Average Percent
Whites Voting
for Black Candidates
22%
Average Percent
Blacks Voting For
White Candidates
23%
36
LICH TM AN ELECTIONS PLUS R E S T OF
GOVERNOR, LABOR A N D CONGRESS (No. 2)
Average Percent Averace Percent
Whites Voting Blacks Voting For
for Black Candidates White Candidates
28% 20%
LICH TM AN ELECTIONS PLUS R E ST OF
GOVERNOR LABOR, CONGRESS AND JUDICIAL
(No. 3)
Average Percent Average Percent
Whites Voting Blacks Voting For
for Black Candidates White Candidates
38% 20%
V. A N A N A L Y S IS OF TH E L IK E L IH O O D OF
B LA C K C A N D ID A TE S P R E V A IL IN G OVER
W H I T E O P P O N E N T S I N D I S T R I C T S OF
VARYIN G P ER C E N TA G E S OF B LA C K VOT
ERS.
I have been asked to perform a statistical analysis, if
possible, that predicts the likelihood of black candidates
prevailing over white opponents, based on the existing
data, as a function of the racial percentage of the elec
tion district. Neither Dr. Lichtman nor Dr. Weber have
done such an analysis.
F irst, I determ ined the frequency distribution for
black candidates winning individual precincts in the
available data sets as a function of the racial composition
of those precincts. From that frequency distribution, I
then calculated the probabilities of black candidates
winning in entire election districts as a function of the
percentage of black registered voters in the district.
These results were graphically plotted for several data
sets. In making these plots, I have used intervals of five
37
percent (for example, a range of 40 percent to 45 per
cent, and then 45 percent to 50 percent, of black regis
tered voters). The results of these probability analyses
are contained after Tab 5.
Generally speaking, the results of this analysis indi
cate that black candidates in the several election data
bases have relatively low chances of winning against
white opponents when the percentage of black regis
tered voters is relatively small. However, the chances of
black success are by no means zero even when the per
centage of black registered voters in a district may be as
low as 25 or 30 percent. Then, the likelihood of a black
candidate winning the election and defeating his white
opponents increases steadily until the percent of black
re g is te re d v o te rs becom es a sign ifican t m ajo rity
(approximately 55-60 percent, although the num ber
varies depending on the specific election database). The
likelihood of an African-American candidate winning
plateaus in the 70-80 percent range, and then increases
slowly as the percentage of black reg istered voters
increases on up to 100 percent.
38
D R . L IC H T M A N ’S P R IM A R Y E L E C T IO N D A T A B A SE
(R E C E IV E D 06/94)
B A L D W IN C O U N T Y
C O U N T Y C O M M IS S IO N E R (1) 1988
IN IT IA L E L E C T IO N
V O T E R R E G IS T R A T IO N , 1988
W H IT E V O T E R B L A C K V O T E R
R E G IS T R A T IO N , 5506 R E G IS T R A T IO N , 2241
5 DATA P O IN T S
E S T IM A T E D P E R C E N T O F
W H IT E V O T E S C A S T F O R
W H IT E C A N D ID A T E (S ): 41%
E S T IM A T E D P E R C E N T O F
B L A C K V O T E S C A S T F O R
B L A C K C A N D ID A T E (S ): 16%
S Q U A R E D C O R R E L A T IO N
(R -SQ )
W H IT E : 0.0122
P -V A L U E M E A S U R E O F
S T A T IS T IC A L
S IG N IF IC A N C E
W H IT E : 0.8595
S Q U A R E D C O R R E L A T IO N
(R-SQ )
B L A C K : 0.0212
P -V A L U E M E A S U R E O F
S T A T IS T IC A L
S IG N IF IC A N C E
B L A C K : 0.8153
ESTIM ATED TURNOUT
W H IT E B L A C K
V O T E R S V O T E R S TO TA L
F O R W H IT E C A N D (S ) 3.5% (193) 6.6% (149) 341
F O R B L A C K C A N D (S ) 5.0% (276) 1.3% (29) 305
E S T IM A T E D TO TA L
T U R N O U T 8.5% (469) 7.9% (178) 646
39
TURNOUT CONFIDENCE INTERVALS
W H IT E V O T E R S B L A C K V O T E R S
F O R W H IT E C A N D (S ) -26% TO 33% -34% TO 47%
F O R B L A C K C A N D (S ) -21% TO 31% -35% TO 38%
D R . L IC H T M A N ’S P R IM A R Y E L E C T IO N D A T A B A SE
(R E C E IV E D 06/94)
B R A N T L E Y C O U N T Y
C O U N T Y C O M M IS S IO N E R (3) 1980
R U N O F F E L E C T IO N
V O T E R R E G IS T R A T IO N , 1980
W H IT E V O T E R B L A C K V O T E R
R E G IS T R A T IO N , 4549 R E G IS T R A T IO N , 294
9 DATA P O IN T S
E S T IM A T E D P E R C E N T O F E S T IM A T E D P E R C E N T O F
W H IT E V O T E S C A S T F O R B L A C K V O T E S C A S T F O R
W H IT E C A N D ID A T E (S ): 56% B L A C K C A N D ID A T E (S ): 453%
S Q U A R E D C O R R E L A T IO N
(R -SQ )
W H IT E : 0.2912
P -V A L U E M E A S U R E O F
S T A T IS T IC A L
S IG N IF IC A N C E
W H IT E : 0.1337
S Q U A R E D C O R R E L A T IO N
(R -SQ )
B L A C K : 0.1024
P -V A L U E M E A S U R E O F
S T A T IS T IC A L
S IG N IF IC A N C E
B L A C K : 0.4013
ESTIM ATED TURNOUT
W H IT E B L A C K
V O T E R S V O T E R S TO TA L
F O R W H IT E C A N D (S ) 41.7% (1896) -69.1% (-203) 1693
F O R B L A C K C A N D (S ) 32.5% (1479) 88.7% (261) 1740
E S T IM A T E D TO TA L
T U R N O U T 74.2% (3375) 19.6% (58) 3433
40
TURNOUT CONFIDENCE INTERVALS
W H IT E V O T E R S B L A C K V O T E R S
F O R W H IT E C A N D (S ) 12% TO 71% -228% TO 89%
F O R B L A C K C A N D (S ) 4% TO 61% -64% TO 241%
D R . L IC H T M A N ’S P R IM A R Y E L E C T IO N D A TA B A SE
(R E C E IV E D 06/94)
B R O O K S C O U N T Y
S C H O O L B O A R D 1986
R U N O F F E L E C T IO N
V O T E R R E G IS T R A T IO N , 1986
W H IT E V O T E R
R E G IS T R A T IO N , 4120
11 DATA P O IN T S
E S T IM A T E D P E R C E N T O F
W H IT E V O T E S C A S T F O R
W H IT E C A N D ID A T E (S):
152%
S Q U A R E D C O R R E L A T IO N
(R -SQ )
W H IT E : 0.104
B L A C K V O T E R
R E G IS T R A T IO N , 2272
E S T IM A T E D P E R C E N T O F
B L A C K V O T E S C A S T F O R
B L A C K C A N D ID A T E (S ):
66%
S Q U A R E D C O R R E L A T IO N
(R -SQ )
B L A C K : 0.8814
P -V A L U E M E A S U R E O F
S T A T IS T IC A L
S IG N IF IC A N C E
W H IT E : 0.3334
P -V A L U E M E A S U R E O F
S T A T IS T IC A L
S IG N IF IC A N C E
B L A C K : 0
41
ESTIM ATED TURNOUT
W H IT E B L A C K
V O T E R S V O T E R S TO TA L
F O R W H IT E C A N D (S ) 9.6% (397) 23.2% (526) 923
F O R B L A C K C A N D (S ) -3.3% (-136) 44.4% (1010) 874
E S T IM A T E D TO TA L
T U R N O U T 6.3% (261) 67.6% (1536) 1797
TURNOUT CONFIDENCE INTERVALS
W H IT E V O T E R S B L A C K V O T E R S
F O R W H IT E C A N D (S ) -8% TO 27% -1% TO 47%
F O R B L A C K C A N D (S ) -11% T O 5% 34% TO 55%
D R . L IC H T M A N ’S P R IM A R Y E L E C T IO N D A T A B A SE
(R E C E IV E D 06/94)
C O L U M B IA C O U N T Y
C O U N T Y C O M M IS S IO N E R (5) 1984
R U N O F F E L E C T IO N
V O T E R R E G IS T R A T IO N , 1983
W H IT E V O T E R B L A C K V O T E R
R E G IS T R A T IO N , 12805 R E G IS T R A T IO N , 1843
12 DA TA P O IN T S
E S T IM A T E D P E R C E N T O F
W H IT E V O T E S C A S T F O R
W H IT E C A N D ID A T E (S ):
59%
S Q U A R E D C O R R E L A T IO N
(R -SQ )
W H IT E : 0.2051
E S T IM A T E D P E R C E N T O F
B L A C K V O T E S C A S T F O R
B L A C K C A N D ID A T E (S ):
70%
S Q U A R E D C O R R E L A T IO N
(R-SQ )
B L A C K : 0.6216
42
P -V A L U E M E A S U R E O F P -V A L U E M E A S U R E O F
S T A T IS T IC A L S T A T IS T IC A L
S IG N IF IC A N C E S IG N IF IC A N C E
W H IT E : 0.1393 B L A C K 0.0023
ESTIM ATED TURNOUT
W H IT E B L A C K
V O T E R S V O T E R S TO TA L
F O R W H IT E C A N D (S ) 15.6% (2001) 26.5% (489) 2491
F O R B L A C K C A N D (S ) 11.0% (1403) 62.3% (1149) 2552
E S T IM A T E D TO TA L
T U R N O U T 26.6% (3405) 88.9% (1638) 5043
TURNOUT CONFIDENCE INTERVALS
W H IT E V O T E R S B L A C K V O T E R S
F O R W H IT E C A N D (S ) 7% TO 24% 13% TO 40%
F O R B L A C K C A N D (S ) -4% TO 26% 38% TO 87%
D R . L IC H T M A N ’S P R IM A R Y E L E C T IO N D A TA B A SE
(R E C E IV E D 06/94)
C O W E T A JU D IC IA L C IR C U IT
P R E S ID E N T 1988
P R E S ID E N T IA L P R E F E R E N C E E L E C T IO N
V O T E R R E G IS T R A T IO N , 1986
W H IT E V O T E R B L A C K V O T E R
R E G IS T R A T IO N , 5948 R E G IS T R A T IO N , 3149
12 DATA P O IN T S
E S T IM A T E D P E R C E N T O F E S T IM A T E D P E R C E N T O F
W H IT E V O T E S C A S T F O R B L A C K V O T E S C A S T F O R
W H IT E C A N D ID A T E (S ): B L A C K C A N D ID A T E (S ):
103% 101%
43
S Q U A R E D C O R R E L A T IO N
(R-SQ )
W H IT E : 0.4238
P -V A L U E M E A S U R E O F
S T A T IS T IC A L
S IG N IF IC A N C E
W H IT E : 0.0219
S Q U A R E D C O R R E L A T IO N
(R -SQ )
B L A C K : 0.9119
P -V A L U E M E A S U R E O F
S T A T IS T IC A L
S IG N IF IC A N C E
B L A C K : 0
ESTIM ATED TURNOUT
W H IT E B L A C K
V O T E R S V O T E R S TO TA L
F O R W H IT E C A N D (S ) 23.3% (1386) -0.3% (-8) 1378
F O R B L A C K C A N D (S ) -0.7% (-39) 42.9% (1351) 1312
E S T IM A T E D TO TA L
T U R N O U T 22.6% (1347) 42.6% (1343) 2690
TURNOUT CONFIDENCE INTERVALS
W H IT E V O T E R S B L A C K V O T E R S
F O R W H IT E C A N D (S ) 11% TO 36% -16% TO 15%
F O R B L A C K C A N D (S ) -7% TO 5% 35% TO 50%
D R . L IC H T M A N ’S P R IM A R Y E L E C T IO N D A T A B A SE
(R E C E IV E D 06/94)
F R A N K L IN C O U N T Y
B O A R D O F E D U C A T IO N 1984
R U N O F F E L E C T IO N
V O T E R R E G IS T R A T IO N , 1984
W H IT E V O T E R B L A C K V O T E R
R E G IS T R A T IO N , 7793 R E G IS T R A T IO N , 429
13 DATA P O IN T S
44
E S T IM A T E D P E R C E N T O F
W H IT E V O T E S C A S T F O R
W H IT E C A N D ID A T E (S ):
49%
E S T IM A T E D P E R C E N T O F
B L A C K V O T E S C A S T F O R
B L A C K C A N D ID A T E (S ):
117%
S Q U A R E D C O R R E L A T IO N S Q U A R E D C O R R E L A T IO N
(R-SQ) (R-SQ )
W H IT E : 0.0315 B L A C K : 0.2361
P -V A L U E M E A S U R E O F P -V A L U E M E A S U R E O F
S T A T IS T IC A L S T A T IS T IC A L
S IG N IF IC A N C E S IG N IF IC A N C E
W H IT E : 0.5617 B L A C K : 0.0923
ESTIM ATED TURNOUT
W H IT E B L A C K
V O T E R S V O T E R S TO TA L
F O R W H IT E C A N D (S ) 24.9% (1941) 1.7% (7) 1948
F O R B L A C K C A N D (S ) 26.5% (2061) -11.8% (-50) 2011
E S T IM A T E D TO TA L
T U R N O U T 51.4% (4002) -10.1% (-43) 3959
TURNOUT CONFIDENCE INTERVALS
W H IT E V O T E R S B L A C K V O T E R S
F O R W H IT E C A N D (S ) 10% TO 40% -97% TO 101%
F O R B L A C K C A N D (S ) 19% TO 34% -65% TO 41%
45
D R . L IC H T M A N ’S P R IM A R Y E L E C T IO N D A T A B A SE
(R E C E IV E D 06/94)
IR W IN C O U N T Y
C O U N T Y C O M M IS S IO N E R (1) 1980
IN IT IA L E L E C T IO N
V O T E R R E G IS T R A T IO N , 1981
W H IT E V O T E R B L A C K V O T E R
R E G IS T R A T IO N , 2909 R E G IS T R A T IO N , 560
10 DATA P O IN T S
E S T IM A T E D P E R C E N T O F
W H IT E V O T E S C A S T F O R
W H IT E C A N D ID A T E (S):
57%
S Q U A R E D C O R R E L A T IO N
(R -SQ )
W H IT E : 0.0604
P -V A L U E M E A S U R E O F
S T A T IS T IC A L
S IG N IF IC A N C E
W H IT E : 0.4937
E S T IM A T E D P E R C E N T O F
B L A C K V O T E S C A S T F O R
B L A C K C A N D ID A T E (S ):
-2677%
S Q U A R E D C O R R E L A T IO N
(R-SQ )
B L A C K : 0.2961
P -V A L U E M E A S U R E O F
S T A T IS T IC A L
S IG N IF IC A N C E
B L A C K : 0.1039
ESTIM ATED TURNOUT
W H IT E B L A C K
V O T E R S V O T E R S TOTA L
F O R W H IT E C A N D (S ) 42.2% (1228) 16.7% (93) 1321
F O R B L A C K C A N D (S ) 32.3% (939) -16.1% (-90) 849
E S T IM A T E D TO TA L
T U R N O U T 74.5% (2167) 0.6% (3) 2170
46
TURNOUT CONFIDENCE INTERVALS
W H IT E V O T E R S B L A C K V O T E R S
F O R W H IT E C A N D (S ) 15% TO 70% -64% TO 97%
F O R B L A C K C A N D (S ) 12% TO 53% -75% TO 43%
D R . L IC H T M A N ’S P R IM A R Y E L E C T IO N D A T A B A SE
(R E C E IV E D 06/94)
IR W IN C O U N T Y
C O U N T Y C O M M IS S IO N E R (1) 1980
R U N O F F E L E C T IO N
V O T E R R E G IS T R A T IO N , 1981
W H IT E V O T E R
R E G IS T R A T IO N , 874
6 DATA P O IN T S
E S T IM A T E D P E R C E N T O F
W H IT E V O T E S C A S T F O R
W H IT E C A N D ID A T E (S ):
52%
S Q U A R E D C O R R E L A T IO N
(R -SQ )
W H IT E : 0.014
P -V A L U E M E A S U R E O F
S T A T IS T IC A L
S IG N IF IC A N C E
W H IT E : 0.8233
B L A C K V O T E R
R E G IS T R A T IO N , 115
E S T IM A T E D P E R C E N T O F
B L A C K V O T E S C A S T F O R
B L A C K C A N D ID A T E (S ):
-32%
S Q U A R E D C O R R E L A T IO N
(R-SQ )
B L A C K : 0.1394
P -V A L U E M E A S U R E O F
S T A T IS T IC A L
S IG N IF IC A N C E
B L A C K : 0.466
47
ESTIM ATED TURNOUT
W H IT E B L A C K
V O T E R S V O T E R S TOTA L
F O R W H IT E C A N D (S ) 37.7% (329) 57.3% (66) 395
F O R B L A C K C A N D (S ) 34.3% (300) -13.9% (-16) 284
E S T IM A T E D TO TA L
T U R N O U T 72.0% (629) 43.5% (50) 679
TURNOUT CONFIDENCE INTERVALS
W H IT E V O T E R S B L A C K V O T E R S
F O R W H IT E C A N D (S ) -29% TO 104%> -171% TO 285%
F O R B L A C K C A N D (S ) -14% TO 83% -179% TO 151%
D R . L IC H T M A N ’S P R IM A R Y E L E C T IO N D A T A B A SE
(R E C E IV E D 06/94)
J E N K I N S C O U N T Y
C O U N T Y C O M M IS S IO N E R (1) 1972
R U N O F F E L E C T IO N
V O T E R R E G IS T R A T IO N , 1972
W H IT E V O T E R B L A C K V O T E R
R E G IS T R A T IO N , 2940 R E G IS T R A T IO N , 1152
7 DATA P O IN T S
E S T IM A T E D P E R C E N T O F
W H IT E V O T E S C A S T F O R
W H IT E C A N D ID A T E (S ):
54%
S Q U A R E D C O R R E L A T IO N
(R-SQ )
W H IT E : 0.0621
P -V A L U E M E A S U R E O F
S T A T IS T IC A L
S IG N IF IC A N C E
W H IT E : 0.5899
E S T IM A T E D P E R C E N T O F
B L A C K V O T E S C A S T F O R
B L A C K C A N D ID A T E (S ):
49%
S Q U A R E D C O R R E L A T IO N
(R -SQ )
B L A C K : 0.1589
P -V A L U E M E A S U R E O F
S T A T IS T IC A L
S IG N IF IC A N C E
B L A C K : 0.3758
48
ESTIM ATED TURNOUT
W H IT E B L A C K
V O T E R S V O T E R S TO TA L
F O R W H IT E C A N D (S ) 24.3% (715) 60.6% (698) 1413
F O R B L A C K C A N D (S ) 20.3% (598) 58.9% (678) 1276
E S T IM A T E D TO TA L
T U R N O U T 44.6% (1312) 119.5% (1377) 2689
TURNOUT CONFIDENCE INTERVALS
W H IT E V O T E R S B L A C K V O T E R S
F O R W H IT E C A N D (S ) -8% TO 57% -16% TO 137%
F O R B L A C K C A N D (S ) -0% TO 41% 11% TO 107%
49
L IC H T M A N DATA (R E V IS E D )
B A L D W IN C O U N T Y
C O U N T Y C O M M IS S IO N E R (1) 1988
IN I T IA L E L E C T IO N
V O T E R R E G IS T R A T IO N , 1988
5 DA TA P O IN T S
E S T IM A T E D P E R C E N T O F E S T IM A T E D P E R C E N T O F
W H IT E V O T E S C A S T F O R B L A C K V O T E S C A S T F O R
W H IT E C A N D ID A T E (S ): 17.7 % B L A C K C A N D ID A T E (S ): -40%
C O N F ID E N C E IN T E R V A L C O N F ID E N C E IN T E R V A L
W H IT E V O T E F O R W H IT E B L A C K V O T E F O R B L A C K
C A N D ID A T E (S ): C A N D ID A T E (S ):
-3.9% TO 39.3 % -69.9% TO -10.1%
S Q U A R E D C O R R E L A T IO N (R-SQ ): 0.92
P -V A L U E M E A S U R E O F S T A T IS T IC A L S IG N IF IC A N C E :
0.01
L IC H T M A N DATA (R E V IS E D )
B R A N T L E Y C O U N T Y
C O U N T Y C O M M IS S IO N E R (3) 1980
R U N O F F E L E C T IO N
V O T E R R E G IS T R A T IO N , 1980
9 DATA P O IN T S
E S T IM A T E D P E R C E N T O F
W H IT E V O T E S C A S T F O R
W H IT E C A N D ID A T E (S ):
56.4%
E S T IM A T E D P E R C E N T O F
B L A C K V O T E S C A S T F O R
B L A C K C A N D ID A T E (S ):
164.9%
C O N F ID E N C E IN T E R V A L
W H IT E V O T E F O R W H IT E
C A N D I D A T E (S):
16.9% TO 95.9%
C O N F ID E N C E IN T E R V A L
B L A C K V O T E F O R B L A C K
C A N D ID A T E (S ):
-47.5% TO 377.3%
50
S Q U A R E D C O R R E L A T IO N (R-SQ ): 0.21
P -V A L U E M E A S U R E O F S T A T IS T IC A L S IG N IF IC A N C E :
0.214
L IC H T M A N DATA (R E V IS E D )
B R O O K S C O U N T Y
S C H O O L B O A R D 1986
R U N O F F E L E C T IO N
V O T E R R E G IS T R A T IO N , 1986
11 DATA P O IN T S
E S T IM A T E D P E R C E N T O F E S T IM A T E D P E R C E N T O F
W H IT E V O T E S C A S T F O R B L A C K V O T E S C A S T F O R
W H IT E C A N D ID A T E (S ): 87.1% B L A C K C A N D ID A T E (S ): 100%
C O N F ID E N C E IN T E R V A L C O N F ID E N C E IN T E R V A L
W H IT E V O T E F O R W H IT E B L A C K V O T E F O R B L A C K
C A N D ID A T E (S ): C A N D ID A T E (S ):
54,3% TO 119.9% 55.7% TO 144.3%
S Q U A R E D C O R R E L A T IO N (R-SQ): 0.51
P -V A L U E M E A S U R E O F S T A T IS T IC A L S IG N IF IC A N C E :
0.014
L IC H T M A N DATA (R E V IS E D )
C O L U M B IA C O U N T Y
C O U N T Y C O M M IS S IO N E R (5) 1984
R U N O F F E L E C T IO N
V O T E R R E G IS T R A T IO N , 1983
12 DATA P O IN T S
E S T IM A T E D P E R C E N T O F E S T IM A T E D P E R C E N T O F
W H IT E V O T E S C A S T F O R B L A C K V O T E S C A S T F O R
W H IT E C A N D ID A T E (S ): 55.7% B L A C K C A N D ID A T E (S ): 82.3%
51
C O N F ID E N C E IN T E R V A L
W H IT E V O T E F O R W H IT E
C A N D ID A T E (S ):
33.6% TO 77.8%
C O N F ID E N C E IN T E R V A L
B L A C K V O T E F O R B L A C K
C A N D ID A T E (S ):
46.9% TO 117.7%
S Q U A R E D C O R R E L A T IO N (R-SQ ): 0.4
P -V A L U E M E A S U R E O F S T A T IS T IC A L S IG N IF IC A N C E :
0.028
L IC H T M A N DATA (R E V IS E D )
IR W IN C O U N T Y
C O U N T Y C O M M IS S IO N E R (1) 1980
IN IT IA L E L E C T IO N
V O T E R R E G IS T R A T IO N , 1981
10 DATA P O IN T S
E S T IM A T E D P E R C E N T O F E S T IM A T E D P E R C E N T O F
W H IT E V O T E S C A S T F O R B L A C K V O T E S C A S T F O R
W H IT E C A N D ID A T E (S ): 56.5% B L A C K C A N D ID A T E (S ): 14.5%
C O N F ID E N C E IN T E R V A L
W H IT E V O T E F O R W H IT E
C A N D ID A T E (S ):
22.8% TO 90.2%
C O N F ID E N C E IN T E R V A L
B L A C K V O T E F O R B L A C K
C A N D ID A T E (S ):
-84.4% TO 113.4%
S Q U A R E D C O R R E L A T IO N (R-SQ): 0.06
P -V A L U E M E A S U R E O F S T A T IS T IC A L S IG N IF IC A N C E :
0.514
L IC H T M A N DATA (R E V IS E D )
IR W IN C O U N T Y
C O U N T Y C O M M IS S IO N E R (1) 1980
R U N O F F E L E C T IO N
V O T E R R E G IS T R A T IO N , 1981
6 DATA P O IN T S
52
E S T IM A T E D P E R C E N T O F E S T IM A T E D P E R C E N T O F
W H IT E V O T E S C A S T F O R B L A C K V O T E S C A S T F O R
W H IT E C A N D ID A T E (S ): 52.1% B L A C K C A N D ID A T E (S ): -6.1%
C O N F ID E N C E IN T E R V A L C O N F ID E N C E IN T E R V A L
W H IT E V O T E F O R W H IT E B L A C K V O T E F O R B L A C K
C A N D ID A T E (S): C A N D ID A T E (S ):
-24.8% TO 129% -268.4% TO 256.2%
S Q U A R E D C O R R E L A T IO N (R-SQ): 0.07
P -V A L U E M E A S U R E O F S T A T IS T IC A L S IG N IF IC A N C E :
0.604
L IC H T M A N DATA (R E V IS E D )
J E N K I N S C O U N T Y
C O U N T Y C O M M IS S IO N E R (1) 1972
R U N O F F E L E C T IO N
V O T E R R E G IS T R A T IO N , 1972
7 DATA P O IN T S
E S T IM A T E D P E R C E N T O F E S T IM A T E D P E R C E N T O F
W H IT E V O T E S C A S T F O R B L A C K V O T E S C A S T F O R
W H IT E C A N D ID A T E (S ): 53.9% B L A C K C A N D ID A T E (S ): 50.9%
C O N F ID E N C E IN T E R V A L C O N F ID E N C E IN T E R V A L
W H IT E V O T E F O R W H IT E B L A C K V O T E F O R B L A C K
C A N D ID A T E (S): C A N D ID A T E (S ):
18.2% TO 89.6% -32.1% TO 133.9%
S Q U A R E D C O R R E L A T IO N (R-SQ): 0
P -V A L U E M E A S U R E O F S T A T IS T IC A L S IG N IF IC A N C E :
0.948
53
D. TESTIMONY OF ALLAN J. LICHTMAN
[Excerpts from Trial Testimony 4 T. Tr. p. 199, line 15
to p. 207, line 22; p. 227, line 17 to p. 228, line 24]
MS. M URPHY: F o r the C o u rt’s re fe ren ce , Dr.
Lichtman has prepared a report which is labeled as DOJ
Exhibit “24” as well as a supplemental report which is
labeled DOJ Exhibit “41”. I would ask that both of those
be admitted into evidence.
Dr. Lichtman will be discussing them here.
HONORABLE JUDGE BOWEN: “24” and “41” are
admitted.
Q (Ms. Murphy) How did you conduct your analysis
of racially polarized voting, Dr. Lichtman?
A I used a two-method standard in the field and
about which I’ve written. One is a ecological regression
analysis which examines all the precincts within a dis
trict. And, the other one is extreme case analysis which
focuses on the most heavily white and the most heavily
black precincts in a d istrict, e ither it be 80% or for
Districts Eleven and Two at the 90% level.
These are the same methodologies utilized by Dr.
Bernard Grofman, the expert for Plaintiffs in Thornburg
versus Gingles, and they are the same methods that
both Dr. Grofman and I used as co-experts in the Los
Angeles County redistricting matter, Garza versus Los
Angeles County, and they are discussed in my books
and articles.
Q Were you able to draw any conclusions concerning
racially polarized voting?
A Yes.
Q And, what were those?
54
A F irst of all, I examined more than 300 elections
statewide in districts throughout the state covering a
great variety of different public offices and spanning
about a 20-year period.
W hat I found was a very common p a tte rn among
those more than 300 elections such that there was the
presence of racially polarized voting. Specifically, I
found that black voters were usually cohesive behind
black candidates running in these elections and I should
say these are all black versus white contests that empir
ically black voters overwhelmingly tended to favor
black candidates in more than 300 black versus white
contests.
And, conversely, white voters usually in overwhelm
ing numbers tended to favor the white candidates com
peting in those elections. And, I found that there was no
p articu la r p a tte rn over tim e e ith e r suggesting an
increase or a diminution of racially polarized voting.
Rather, you have the presence of racially polarized vot
ing throughout the period.
This is confirm ed both by ecological reg ression
analysis and the analysis that spotlights the heavily
black and heavily white precincts. It is also sustained by
the standard checks that Dr. Grofman and I employ and
that’s discussed in the literature for reliability of these
results.
Q W ere th e re any exceptions to the p a tte rn of
racially polarized voting that you found?
A There are always some exceptions. In social sci
ence we do not look for uniform patterns, we look for
unusual patterns. And, perhaps the most notable excep
tion would be judicial elections with black incumbents.
But, the exceptions are few.
55
Q Are the results of this analysis you’ve been dis
cussing reported in the appendices to your report?
A Yes, Appendix 1 and 2, I report the ecological
regression and the extreme case analysis, respectively.
Q And, are most of the elections reported in the
appendices, were they 1990 or prior? Were most of those
elections for 1990 going back to 1970?
A Yes, the great bulk of the analyses I did was done
when only elections through 1990 was available. I t
should also be noted th a t these elections focus on
Democratic primary and runoff elections.
And, I believe those to be the significant elections in
the m atter before us as well because it is the Demo
cratic primaries and runoffs that essentially determine
who is going to be elected from the E leven th and
Second Congressional Districts.
Q In your analysis of racially polarized voting, did
you also do some analyses that focus specifically on the
Eleventh and the Second Congressional District?
A Yes, I did.
Q And, what was that analysis?
A What I was able to do for those districts specifi
cally was examine elections conducted on a statewide
basis, but isolated within the boundaries of Districts
Eleven and Two. I was able to analyze some eight such
elections for District Eleven, including of course, the
1992 primary and runoff election in those districts.
These are all black versus white primary or runoff
elections in District Eleven and I was able to examine
some seven such statewide black versus white elections
specifically within the boundaries of D istrict Two as
well.
56
Q And, did you draw any conclusions from that analy
sis?
A Yes. What I found that when you spotlighted elec
tions within the Eleventh and Second Congressional
D istrict, was tha t voting was indeed polarized along
racial lines with a very substantial majority of black vot
ers usually voting for the black candidate or candidates in
these elections and likewise a very substantial majority
of white voters usually voting for the white candidate or
candidates in these elections and that was true of both
the Eleventh and Second Congressional Districts.
Q Were there any exceptions to that pattern that you
just discussed?
A Yes, there were.
Q What were they?
A When you look at the Democratic Prim ary and
Democratic Runoff elections w ithin the confines of
District Eleven, you find essentially that there are two
exceptions, both elections involving multiple black can
didates and both of them primary elections only.
In the D istrict Eleven prim ary for Congress, the
percent of white voters voting for black candidates was
a majority, 55%, and in the 1992 Democratic Primary for
Labor Commissioner, the percent of white voters voting
for black candidates was 45%.
Within District Two there is one exception and that
exception is the 1992 Democratic Prim ary for Labor
Commissioner in which 52% of white voters voted for
black—the two black candidates in the initial primary.
Q For each of those three elections you just noted,
did you also examine the Democratic Primary runoffs?
57
A Yes, each of those elections went to a runoff and
you found when you went to the runoff elections polar
ization substantially increased and returned to the more
usual patterns of substantial majorities of black voters
voting for the single black candidate and substantial
majorities of white voters voting for the single white
candidate.
In all cases, e ither blacks for blacks or whites for
whites, it was in excess of 70%.
Q Are the results of that analysis you’ve just been
discussing in D istricts E leven and Two reported in
Tables 1 through 4 of your report?
A Yes.
Q Do you have a definition, Dr. Lichtman, of politi
cally significant racially polarized voting?
A Yes.
Q And, what would that be?
A Racially polarized voting exists in a kind of a raw
sense when blacks and whites differ in their choice of
candidates. I t becomes politically significant when dif
ferences in the choice of candidates has an impact upon
the ability of black voters, or generally minority voters,
to elect candidates of their choice.
Q And, did you apply tha t definition here in this
analysis that you did?
A Yes.
Q And, what was your conclusion?
A As follows. Given the substantial differences in the
usual patterns of black and white voting, there is cer
tainly the potential for voting to become politically sig
nificant. That is, there is certainly the potential for
58
white-black voting, given black cohesion to have an
impact on the ability of black voters to elect candidates
of their choice.
However, in—particularly in the Eleventh District,
but in the Eleventh and the Second District, there turns
out to be a sufficient concentration of minority voters
that both majority black districts such that even given
the polarization the black voters are able to elect candi
dates of their choice. So, specific to these districts politi
cally significant racially polarized voting does not emerge.
Q Do you have an opinion as to whether if the black
percentages, the black percentage of population in these
districts was reduced, w hether voting might become
politically significantly racially polarized?
MR. PARKS: Your Honor, if I might. I think tha t
there’s a pretty significant difference in the percentage
of those districts and I think we need to ask that ques
tion by district.
HONORABLE JUDGE BOWEN: Do you mind sepa
rating it?
A I certainly don’t.
HONORABLE JUDGE BOWEN: All right.
A F irst of all, for the Eleventh District, you’re deal
ing with a district that’s around 60% in its voting age
population. Under conditions of racially polarized vot
ing, to the extent you reduce the percentages of minori
ties in the district, you are going to reduce the opportu
nities for minority voters to elect candidates of choice.
So, as you move down from approx im ate ly 60%
towards the 50% level and above, you are obviously
going to increase the potential for politically significant
racially polarized voting. And, if this were to become a
59
majority white or perhaps even a near-majority white
district, it is my view you would have politically signifi
cant racially polarized voting.
Obviously, in the Second D istrict, because you’re
dealing with a much lower initial percentage, you have a
smaller margin. And, in this case, smaller reductions
wouldn’t take much to transform that district into a
majority white district and you would again have the
potential for politically significant racially polarized vot
ing to operate.
Q (Ms. Murphy) Moving on to the second issue in
your report, Dr. Lichtman, did you do some examination
of socioeconomic differences and political consequences?
A Yes, the second thing I looked at was the extent of
differences in the socioeconomic standing of whites and
blacks within the State of Georgia and then at a ques
tion of whether such socioeconomic differences, if they
emerge, have political consequences for the ability of
blacks to participate fully in the political process and
elect candidates of their choice.
Q How did you analyze that question?
A First, with respect to socioeconomic differences it’s
quite sim ple. I looked at the 1990 U.S. Census of
Housing and Population. 1 looked at standard measures
of income occupation and education and, indeed, would
find on such standard measures that there are substan
tial differences between blacks and whites such that
blacks and whites—such that whites do have a substan
tially higher socioeconomic status.
Secondly, I looked at the political consequences of
such differences in socioeconomic status. And, the politi
cal consequences are that differences in socioeconomic
status affect the political resources, particularly under
60
conditions of racially polarized voting available to
groups, such things as campaign finance, facilities, well-
educated candidates, transportation and communication.
I t ’s also well known that socioeconomic differences
can have an impact on turnout, the participation of the
advantaged versus the disadvantaged group. I t’s always
a barrier, but sometimes the disadvantaged group is
able to overcome that barrier. That was not the case,
however, in the State of Georgia.
* * *
Q (Ms. Murphy) What other analysis is included or
reported in your supplemental report, Dr. Lichtman?
A Let’s see. The second thing I did was—and these, I
believe are the only elections I looked at that were not
my primary focus elections, Democratic Primaries and
particular runoffs. I looked at some non-partisan judicial
elections conducted statewide in 1990 and 1992 that you
could look at in Districts Eleven and Two.
Q And, why did you analyze those judicial elections,
Dr. Lichtman?
A Well, there have been some judicial elections in my
early appendix because I included some results from a
previous report I had done and but more specifically Dr.
Weber had analyzed these elections and given some
opinions about it and I thought it prudent, given that I
had time between the two reports, to do the analysis
myself.
Q And, what conclusions were you able to draw from
analyzing these elections?
A Well, I think it’s quite clear that the patterns in
these judicial elections are different than the over
whelming patterns in the other elections. And, in my
61
view, as I said, these are the only elections that weren’t
Democratic Primaries or runoffs, they’re a very limited
utility. They are non-partisan.
And, in addition, each of them includes an appointed
black incumbent and I’ve now had experience analyzing
judicial elections throughout the south in numerous
states and my experience is that judicial elections with
black incumbents are quite different from other elec
tions. They’re quite different from non-judicial elections
and they’re even different from open-seat black versus
white judicial elections. So, while I’ve included them in
here for completeness, I believe they’re of minimal rele
vance and certainly do not affect any prior conclusions.
62
E. REPORT [BY DR. ALLAN J. LICHTMAN] ON
ISSUES RELATING TO GEORGIA
CONGRESSIONAL DISTRICTS
[Defendants’ Exh. 24, at p. 3 line 7-p. 20, line 1;
pp. 30-31, & Charts from Appendix 1]
METHODOLOGY AND DATA
The main data base for this study consists of precinct-
by-precinct election returns for numerous black versus
white elections held statewide and within various juris
dictions in the sta te of Georgia. The data base also
includes official, state of Georgia precinct-by-precinct
registration data, with black and white registrants sepa
rately identified. The U.S. Department of Justice and
the state of Georgia provided the election returns, regis
tration data, and racial identification of candidates.
The analysis of racially polarized voting focuses on
Democratic primary and runoff elections. Districts 11
and 2 are sufficiently Democratic that a primary victory
is virtually tantam ount to election. Their Democratic
predom inance is docum ented by overw helm ing
Democratic victories in statew ide general elections
within the precincts of Districts 11 and 2 as well as by
the strongly Democratic results of the 1992 general elec
tions for Congress in Districts 11 and 2.1
S tandard s ta tis tica l m ethods w ere employed to
explore patterns of racially polarized voting by blacks
and w hites as well as p a tte rn s of vo ter tu rnou t by
blacks and whites. An examination of racially polarized
!A s p a r t o f th e d a ta i t p re p a re d fo r th e re d is tr ic t in g p ro c ess th e
s ta te “reco m p iled ” g e n e ra l e lec tion r e tu r n s fo r th e p re c in c ts of p ro
p o s e d D i s t r i c t s 11 a n d 2. T h is i n f o r m a t i o n is c o n ta in e d in
R e a p p o r t io n m e n t S e rv ic e s O ffice , G e o rg ia G e n e ra l A sse m b ly ,
‘“C onference R e p o r t ,” M arch 31 ,1992.
63
voting considers both the cohesion of the black elec
torate and bloc voting by the white electorate.
The analysis of black voter cohesion and bloc voting
by w h ites follows p rocedures recognized by the
Supreme Court in Thornburg v. Gingles, 478 U.S. 30
(1986) and applied by the Court to single-member dis
tricts plans in Quilter v. Voinovich 113 S.Ct 1149 (1993).
Appropriate procedures for analyzing these issues are
also discussed in the social science litera tu re.2
The voting behavior of whites and blacks in Georgia is
estimated by comparing the racial composition of the
various voting precincts to the division of the vote
am ong com peting cand idates in each prec inct.
Ecological regression, the standard method for inferring
the behavior of population groups from data collected
for aggregate units, was used to estim ate the voting
behavior of blacks and whites.
The ecological regression procedure for analyzing the
behavior of v o te r groups is se t fo rth in my book,
Ecological Inference (Sage Series on Q uantitative
2S e e , f o r e x a m p le , C h a n d le r D a v id s o n , e d . , M inority Vote
D ilu tion (W a s h in g to n , D C : H o w a rd U n iv e r s i t y P r e s s , 1984);
R ic h a r d E n g s t r o m , “Q u a n t i t a t i v e E v id e n c e in V o te D i lu t io n
L it ig a tio n : P o lit ic a l P a r t ic ip a t io n and P o la r iz e d V o tin g ,” Urban
Lawyer (1985); J a m e s L o ew en and B e rn a rd G rofm an , “C om m en t:
R e c e n t D e v e lo p m e n t s in M e th o d s U s e d in V o t in g R i g h t s
L itig a tio n ,” Urban Lawyer (1989); A llan J . L ich tm an , “P a s s in g th e
T est: E co logical R e g re s s io n in th e L os A n g e le s C o u n ty C ase and
B e y o n d ,” E v a lu a tio n R eview (1 9 9 1 ); B e r n a r d G r o f m a n a n d
C h a n d le r D a v id so n , ed s ., Controversies in M inority Voting: The
Voting Rights Act in Perspective (1992); B e rn a rd G ro fm an , L isa
H and ley , and R ich a rd G. N iem i, Minority Representation and the
Quest fo r Voting E quality (C a m b rid g e : C a m b r id g e U n iv e r s i ty
P re s s , 1992); A llan J . L ich tm an and J . G era ld H e b e r t , “A G enera l
T h e o ry o f V ote D ilu tion ,” La Raza Law Journal (1993).
64
Applications in Social Science, 1978: with Laura Irwin
Langbein) and analyzed, in depth, in my December, 1991
article in Evaluation Review.
Black and white voting can also be examined through
a technique termed extreme case analysis that examines
the actual choices of voters in the most heavily black
and the most heavily white precincts in a jurisdiction.
The extreme case results will not correspond exactly
with the results of ecological regression analysis. The
80%+ and 90%+ black and white precincts used in this
study obviously include no mid-range precincts and
include some voters of the other race.3 Unlike ecological
regression, extreme case analysis involves no inferential
procedures. It simply tallies the votes actually cast for
black and white candidates in the heavily black and
heavily white precincts.
The primary techniques of ecological regression and
extreme case analysis are supplemented by the exami
nation of squared correlation coefficients and measures
of statistical significance. The squared correlation coeffi
cient (R2) measures the degree to which voting for com
peting candidates can be predicted by knowledge of the
racial composition of precincts alone. The value of R2
varies from 0 to 1.0. Although no values a rb itrarily
define the distinction between “high” and “low,” politi
cal analysts often rely on coefficients with values of
about .25 and consider values of about .5 or greater as
indicative of a strong association between variables.
3F o r ind iv idua l c o u n ty a n a ly se s th e 80 p e rc e n t cu to ff is u sed fo r
e x tr e m e ca se a n a ly s is . F o r th e p o p u lo u s c o n g re ss io n a l d is t r ic ts ,
w ith m o re heav ily w h ite an d h eav ily b lack p re c in c ts , th e 90 p e rc e n t
cu to ff is u sed .
65
Conventionally, social scientists accept as statistically
significant results of either .05 (corresponding to a 5 in
100 probability of obtaining the results under the chance
hypothesis) or the more stringent .01 (corresponding to
a 1 in 100 probability of obtaining the results under the
chance hypothesis). Virtually all findings of this study
are statistically significant at the .01 level.
Election results for the eleventh and second congres
sional districts in Georgia included precincts tha t are
split between two congressional districts. There are
about 20 split precincts of some 250 precincts in District
11 and about 30 split precincts of some 270 precincts in
District 2. Split precincts were treated in two ways.
First, the analysis excluded the split precincts entirely,
incorporating only the precincts entirely within District
11 and D istrict 2 respectively. Unless the voting of
whites and blacks in the split precincts differed funda
mentally from the whites and blacks in the undivided
precincts, this procedure should have little or no effect
on the results of the statistical analysis.
Second, as a check on the first analysis, the voter reg
istration in split precincts was proportionally allocated
to each congressional district studied. The proportion of
registration in a split precinct to be statistically allo
cated to a congressional d istrict was determined by
examining the ratio of the voter turnout in the split
precinct to the registered voters in that precinct.4 The
4I f th e r a t io o f v o te r tu r n o u t to r e g is t r a t io n fo r a g iv e n s p l i t
p re c in c t w as equa l to g r e a te r th a n th e m ean ra tio fo r all u n d iv ided
p re c in c ts , th e e n t i r e ty o f th e re g is tra t io n in th e sp lit p re c in c t w as
s ta tis tic a lly alloca ted to th e cong ressional d is tr ic t. I f th a t ra tio fo r a
g iv e n s p l i t p re c in c t w a s b e lo w th e m e a n r a t io fo r a ll u n d iv id e d
p re c in c ts , on ly a p ro p o r tio n of th e re g is tra t io n in th e sp lit p re c in c t
w as s ta tis tic a lly alloca ted to th e cong ressional d is tr ic t. If, fo r ex am
ple, th e ra tio o f tu rn o u t to re g is tra t io n w as h a lf th a t o f th e m ean
r a t io fo r u n d iv id e d p re c in c ts , o n e -h a lf o f th e r e g is t r a t io n in th e
p re c in c t w as alloca ted to th e cong ressional d is tr ic t.
66
results from the analysis using proportionally allocated
sp lit precincts w ere nearly identical to the resu lts
ob tained from th e analysis th a t excluded sp lit
precincts.
RESULTS I: RACIALLY POLARIZED VOTING
The inquiry into racially polarized voting includes
four levels of analysis: counties throughout the state,
counties within the eleventh and second congressional
d istric ts, statew ide elections partitioned within the
boundaries of congressional districts 11 and 2, and 1992
congressional elections within these districts.
Ecological regression analysis of elections in counties
throughout the state shows a strong pattern of racially
polarized voting for about a 20 year period through
1990, involving a variety of public offices. Ecological
regression analysis of more than 150 elections, reported
in Appendix 1, shows high levels of black cohesion as
black voters usually voted overwhelmingly for black
candidates. Similarly, the ecological regression results in
Appendix 1 show high levels of white bloc voting as
whites voters usually voted overwhelmingly against
black candidates and for their white competitors.
The findings of ecological regression analysis are cor
roborated by the actual results of elections in precincts
that are 80 percent or more black or white, respectively,
in th e ir reg is te red voters. Additional elections are
included in the extrem e case analysis because some
counties with heavily white precincts did not include
sufficiently concentrated populations of black regis-
67
trants for ecological regression analysis.5 The results of
extrem e case analysis, reported in Appendix 2 show
that black candidates almost always garnered majorities
in heavily black precincts, but almost always failed to do
so in heavily white precincts.
The ecological regression and extreme case analyses
conducted throughout the state show no diminution of
racially polarized voting over time. Patterns of racially
polarized voting remain strong throughout the period
studied.
Similarly, patterns of racially polarized voting are
strong in the particular counties included within con
gressional districts 11 and 2. The bulk of the counties
included within districts 11 and 2 are represented in the
study. The following counties included in Appendices 1
or 2 of this study are entirely or in part within District
11: Baldwin, Burke, Chatham, DeKalb, Effingham,
Greene, Hancock, Henry, Jasper, Jefferson, Jenkins,
Putnam , Richmond, Twiggs, W arren, W ashington,
Wilkes, and Wilkinson. The following counties included
in Appendices 1 or 2 are en tirely or in p a rt within
D is tr ic t 2: Baker, Bibb, B rooks, Calhoun, Clay,
Crawford, Decatur, Dooly, Dougherty, Early, Grady,
Houston, Macon, Marion, Meriwether, Muscogee, Peach,
Quitman, Randolph, Seminole, Stewart, Talbot, Taylor,
Terrell, Thomas, and Webster,
The finding of racially polarized voting is likewise
strong when results of statewide black versus white
elections are isolated for the precincts of congressional
districts 11 and 2 respectively. The results of ecological
5T h e re a r e a lso som e co u n tie s fo r w h ic h eco log ical r e g re s s io n
w as p e rfo rm ed th a t d id n o t include 80% + b lack o r w h ite p rec in c ts .
T h e s e c o u n tie s a r e n o t in c lu d e d in t h e e x t r e m e c a se ta b le s in
A p p e n d ix 2.
68
regression and extrem e case analyses for statewide
elections within D istrict 11 are reported in Table 1
below. In all six elections, Table 1 shows that at least 65
percent of black voters were cohesive behind the black
candidates. In 5 of 6 elections, at least 89 percent of
black voters voted for the black candidates.
With respect to white bloc voting, in 5 of 6 elections,
an overwhelming majority of white voters bloc voted
against the black candidates. In these five elections, no
more than 26 percent of white voters supported black
candidates. This means that at least 74 percent of white
voters bloc voted for white candidates. The only excep
tion to this pattern is the 1992 democratic primary for
Labor Commissioner, which included one white and two
black candidates, one of whom was an appointed black
incumbent. In tha t election only a small m ajority of
white voters voted against the black candidates. In
the runoff election for Labor Commissioner, however,
Table 1 d iscloses th a t po lariza tion sign ifican tly
increased; only 26 percent of white voters voted for the
black incumbent candidate, whereas 92 percent of black
voters voted for the black candidate.
The results of ecological regression and extreme case
analyses for statewide elections within District 2 are
reported in Table 2 below.6 In all five elections, Table 2
shows that at least 77 percent of black voters were cohe
sive behind the black candidates. In 4 of 5 elections, an
overw helm ing m ajority of w hite vo ters bloc voted
against the black candidates. In these four elections, no
more than 21 percent of white voters supported black
candidates. This means that at least 79 percent of white
voters bloc voted for white candidates. Again, the only
S u f f ic ie n t d a ta w as n o t av a ilab le to include th e 1984 p re s id e n
t ia l p re fe re n c e p r im a ry . T h u s fiv e r a th e r th a n s ix e le c tio n s a r e
re p o r te d in T able 2.
69
TABLE 1: ECOLOGICAL REGRESSION AND EXTREME
CASE ANALYSIS:
POLARIZATION IN STATEWIDE GEORGIA ELECTONS
PRECINCTS WITHIN BOUNDARIES OF CONGRESSIONAL
DISTRICT 11
ELECTION % BLACK
VOTERS
VOTING
FOR
BLACK
CANDS
% WHITE
VOTERS
VOTING
FOR
BL4CK
CANDS
SQ.
CORR.
COEFF
R2
VOTE FOR
BLACK
CANDS IN
90%+
BLACK
PRECS
VOTE FOR
BLACK
CANDS IN
90%+
WHITE
PRECS
1984 DEM PRIM
PRES. PREF.
BLACK CAND:
JACKSON
WHITE CANDS:
MONDALE
HART
6 OTHERS
65% 0% .76* 66% 7%
1988 DEM PRIM
PRES. PREF
BLACK CAND:
JACKSON
WHITE CANDS:
DUKAKIS
GORE
4 OTHERS
95% 0% .86* 94% 12%
1990 DEM PRIM
GOVERNOR
BLACKCAND:
YOUNG
WHITE CANDS:
MILLER
BARNES
MADDOX
MCDONALD
89% 2% .84* 87% 12%
1990 DEM
RUNOFF
GOVERNOR
YOUNG
MILLER
93% 12% .80* 91% 21%
70
T A B L E L C O N T ’D: E C O L O G IC A L R E G R E S S IO N A N D
E X T R E M E C A S E A N A L Y S IS :
P O L A R IZ A T IO N IN S T A T E W ID E G E O R G IA E L E C T IO N S
P R E C IN C T S W IT H IN B O U N D A R IE S O F C O N G R E S S IO N A L
D IS T R IC T 11
ELECTION % BLACK
VOTERS
VOTING
FOR
BLACK
CANDS
% WHITE
VOTERS
VOTING
FOR
BLACK
CANDS
SQ.
CORR.
COEFF
R2
VOTE FOR
BLACK
CANDS IN
90%+
BLACK
PRECS
VOTE FOR
BLACK
CANDS IN
90%+
WHITE
PRECS
1992 DEM PRIM
LABOR COMM
BLACK CANDS:
JOHNSON
SCOTT
WHITE CAND:
POYTHRESS
96% 45% .77 93% 47%
1992 DEM
RUNOFF LABOR
COMM
SCOTT
POYTHRESS
92% 26% .79* 89% 25%
T H E A N A L Y S IS IN C L U D E S O N L Y P R E C I N C T S
E N T I R E L Y W I T H I N E A C H D IS T R IC T . V IR T U A L L Y
ID E N T IC A L R E S U L T S A R E O B T A IN E D W H E N S P L IT
P R E C I N C T S A R E I N C L U D E D A N D P R O P O R T I O N
A L L Y A D J U S T E D F O R T U R N O U T . T H E R E A R E
A B O U T 20 S P L I T P R E C I N C T S O F S O M E 250 IN D IS
T R IC T 11.
* R E S U L T S S T A T IS T IC A L L Y S I G N I F I C A N T A T
.0001 L E V E L .
71
exception to this pattern is the 1992 Democratic pri
mary for Labor Commissioner. In that election a small
majority of white voters voted for the two black candi
dates. This means that at least 79 percent of white vot
ers bloc voted for white candidates. Again, the only
exception to this pattern is the 1992 Democratic pri
mary for Labor Commissioner. In that election a small
majority of white voters voted for the two black candi
dates. In the runoff election for Labor Commissioner,
however, Table 2 discloses th a t polarization greatly
increased, as only 21 percent of white voters voted for
the black incumbent candidate, whereas 77 percent of
black voters voted for the black candidate.
Finally, the black versus white Democratic primary
and runoff elections for Congress in districts 11 and 2 are
also polarized along racial lines. According to results
reported in Table 3, the black electorate is cohesive in all
four elections, as 85 percent or more of black voters sup
ported black candidates. The results also show white bloc
voting against black candidates. In 3 of 4 elections white
support for black candidates was 26 percent or less. This
means that 74 percent or more of white voters in these
elections voted for white candidates. The only exception
to the pa ttern of white bloc voting occurred in the
District 11 primary, where 4 black candidates competed
against one white candidate. In the District 11 runoff
election, however, polarization increased substantially, as
the black candidate garnered only 23 percent of the
white vote, compared to 97 percent of the black vote.
Candidate-by-candidate analysis of the District 11 pri
mary, reported in Table 4, shows that black candidate
McKinney, one of four black candidates competing in
that election, garnered 20 percent support from white
voters. This placed her second among white voters to
white candidate DeLoach, who garnered 45 percent
72
TABLE 3: ECOLOGICAL REGRESSION AND EXTREME
CASE ANALYSIS:
POLARIZATION IN CONGRESSIONAL ELECTIONS
1992 DEMOCRATIC PRIMARY & RUNOFF
CONGRESSIONAL ELECTIONS
GEORGIA CONGRESSIONAL DISTRICT 11 RESULTS
BASED ON VOTER REGISTRATION DATA
ELECTION % BLACK
VOTERS
VOTING
FOR
BLACK
CANDS
% WHITE
VOTERS
VOTING
FOR
BLACK
CANDS
SQ.
CORR.
COEFF
R2
VOTE FOR
BLACK
CANDS IN
90%+
BLACK
PRECS
VOTE FOR
BLACK
CANDS IN
90%+
WHITE
PRECS
DIST11 PRIM.
BLACK CANDS:
MCKINNEY,
WALKER
THURMOND
LOCKHART
WHITE CAND:
DELOACH
94% 55% .27* 87% 63%
DIST 11 RUN.
MCKINNEY
DELOACH
97% 23% .70* 89% 23%
73
T A B L E 3 C O N T ’D: E C O L O G IC A L A N D E X T R E M E C A S E
A N A L Y S IS :
P O L A R IZ A T IO N IN S T A T E W ID E G E O R G IA E L E C T IO N S
1992 D E M O C R A T IC P R IM A R Y & R U N O F F
C O N G R E S S IO N A L E L E C T IO N S
G E O R G IA C O N G R E S S IO N A L D IS T R IC T 2 R E S U L T S B A S E D
O N V O T E R R E G IS T R A T IO N DATA
ELECTION % BLACK
VOTERS
VOTING
FOR
BLACK
CANDS
% WHITE
VOTERS
VOTING
FOR
BLACK
CANDS
SQ.
CORR.
COEFF
R2
VOTE FOR
BLACK
CANDSIN
90%+
BLACK
PRECS
VOTE FOR
BLACK
CANDSIN
90%+
WHITE
PRECS
DIST 2 PRIM.
BLACK CANDS:
BISHOP
CUMMINGS
EDWARDS
KAIGLER
WHITE CANDS:
HATCHER
WHIGHAM
88% 26% .51* 81% 35%
DIST 2 RUN
BISHOP
HATCHER
85% 18% .48* 75% 29%
T H E A N A L Y S IS I N C L U D E S O N L Y P R E C I N C T S
E N T I R E L Y W I T H I N E A C H D IS T R IC T . V IR T U A L L Y
ID E N T IC A L R E S U L T S A R E O B T A IN E D W H E N S P L IT
P R E C IN C T S A R E IN C L U D E D A N D P R O P O R T IO N A L L Y
A D J U S T E D F O R T U R N O U T . T H E R E A R E A B O U T 20
S P L IT P R E C IN C T S O F SO M E 250 IN D IS T R IC T 11 A N D
A B O U T 30 S P L I T P R E C I N C T S O F S O M E 270 IN D I S
T R IC T 2.
^ R E S U L T S S T A T IS T IC A L L Y S I G N I F I C A N T A T
.0001 L E V E L .
74
T A B L E 4: E C O L O G IC A L R E G R E S S IO N E S T IM A T E S
V O T E S F O R IN D IV ID U A L C A N D ID A T E S
1992 D E M O C R A T IC C O N G R E S S IO N A L P R IM A R Y
D IS T R IC T 11
CANDIDATE % BLACK VOTERS
VOTING FOR
CANDIDATE
% WHITE VOTERS
VOTING FOR
CANDIDATE
MCKINNEY (B) 42% 20%
WALKER (B) 34% 11%
THURMOND (B) 14% 17%
LOCKHART(B) 4% 7%
DELOACH (W) 6% 45%
T H E A N A L Y S IS I N C L U D E S O N L Y P R E C I N C T S
E N T IR E L Y W IT H IN D IS T R IC T 11. V IR T U A L L Y ID E N
T I C A L R E S U L T S A R E O B T A I N E D W H E N S P L I T
P R E C I N C T S A R E I N C L U D E D A N D P R O P O R T I O N
A L L Y A D J U S T E D F O R T U R N O U T . T H E R E A R E
A B O U T 20 S P L I T P R E C I N C T S O F S O M E 250 IN D I S
T R IC T 11.
75
T A B L E 10: E C O L O G IC A L R E G R E S S IO N A N D E X T R E M E
C A S E A N A L Y S IS ; P O L A R IZ A T IO N IN C O N G R E S S IO N A L
E L E C T IO N S
1992 D E M O C R A T IC P R IM A R Y & R U N O F F
C O N G R E S S IO N A L E L E C T IO N S
G E O R G IA C O N G R E S S IO N A L D IS T R IC T 11
R E S U L T S B A S E D O N V O T IN G A G E P O P U L A T IO N DATA
ELECTION % BLACK
VOTERS
VOTING
FOR
BLACK
CANDS
% WHITE
VOTERS
VOTING
FOR
BLACK
CANDS
SQ.
CORR.
COEFF
R2
VOTE FOR
BLACK
CANDS IN
90%+
BLACK
PRECS
VOTE FOR
BLACK
CANDS IN
90%+
WHITE
PRECS
DIST11 PRIM.
BLACK CANDS:
MCKINNEY,
WALKER
THURMOND
LOCKHART WHITE
CAND: DELOACH
94% 54% .23* 85% 63%
DIST 11 RUN.
MCKINNEY
DELOACH
96% 22% .64* 86% 24%
76
T A B L E 10 C O N T ’D: E C O L O G IC A L R E G R E S S IO N A N D
E X T R E M E C A S E A N A L Y S IS ; C O N G R E S S IO N A L
E L E C T IO N S
1992 D E M O C R A T IC P R IM A R Y & R U N O F F
C O N G R E S S IO N A L E L E C T IO N S
G E O R G IA C O N G R E S S IO N A L D IS T R IC T 2
ELECTION % BLACK % WHITE SQ. VOTE FOR VOTE FOR
VOTERS VOTERS CORR. BLACK BLACK
VOTING VOTING COEFF CANDS IN CANDS IN
FOR FOR R2 90%+ 90%+
BLACK BLACK BLACK WHITE
CANDS CANDS PRECS PRECS
DIST 2 PRIM.
BLACK CANDS:
BISHOP
CUMMINGS
EDWARDS
KAIGLER
WHITE CANDS:
HATCHER
WHIGHAM
38% 27% .44* 81% 35%
DIST 2 RUN.
BISHOP
HATCHER
84% 20% .41* 76% 33%
T H E A N A L Y S IS I N C L U D E S O N L Y P R E C I N C T S
E N T I R E L Y W I T H IN E A C H D IS T R IC T . V O T IN G A G E
P O P U L A T IO N DATA W A S N O T A V A IL A B L E F O R T H E
P R E C IN C T S O F R IC H M O N D C O U N T Y IN D IS T R IC T 11
A N D T H E P R E C IN C T S O F M A C O N C O U N T Y IN D IS -
T R IC T 2.
* R E S U L T S S T A T IS T IC A L L Y S I G N I F I C A N T AT
.0001 L E V E L .
77
support from white voters. McKinney’s white support
increased from 20 percent in the primary to only 23 per
cent in the runoff, as white voters coalesced behind the
white candidate. DeLoaeh increased his white support
from 45 percent in the primary to 77 percent in the
runoff election.
RESULTS II: SOCIOECONOMIC DIFFEREN CES
AND POLITICAL CONSEQUENCES
Examination of data from the 1990 U.S. Census o f
Housing and Population dem onstrates clearly th a t
white residents of the state of Georgia have a substan
tially higher socioeconomic status than black residents
of the state of Georgia. Differences between whites and
blacks are found for standard measures of income, occu
pation, and education.
Socioeconomic d ifferences r e s tr ic t th e re la tiv e
resources available to a politically cohesive minority
group such as blacks in Georgia. Such resources include
campaign finance and facilities; a recruitm ent base of
well-educated candidates; and access to transportation
and communication. Lower socioeconomic standing also
constitutes a barrier to the turnout of voters as demon
strated by social science studies.7 Ecological regression
analysis of voter turnout indicates that black residents
of Georgia have not been able to overcome this barrier.
Participation measures generally focus on general elec-
7S e e , f o r e x a m p le , R a y m o n d E . W o lf in g e r a n d S te v e n J .
R o sen sto n e , Who Votes'! (N ew H aven : Y ale U n iv e rs ity P re s s , 1980);
S te v e n J . R o s e n s to n e a n d J o h n M a rk H a n s e n , M obilization,
Participation, and Democracy in America (N ew York: M acm illan,
1993); J a n E . L e ig h le y a n d J o n a th a n N a g le r , “ I n d iv id u a l a n d
S y stem ic In fluences on T urnou t: W ho V otes? 1984,” The Journal of
Politics 52 (1992).
78
tions in which all voters participate and on top-of-the-
ticket contests. Ecological regression analysis of the
1988 presidential election indicates that 62 percent of
white registrants voted, compared to 47 percent of black
registrants. Similarly, ecological regression analysis of
the 1992 presidential election indicates that 72 percent
of white registrants voted, compared to 58 percent of
black registrants. These differences in the turnout of
reg istered w hites and blacks, moreover, understate
overall white/black tu rnou t differences because the
white voter registration rate is higher than the black
voter registration rate. According to the 1990 Census,
about 25 percent of Georgia’s voting age population was
black. But only about 22 percent of registered voters in
1988 and 1992 were black, according to state of Georgia
registration statistics.
79
APPENDIX 1
R A C IA L P O L A R IZ A T IO N IN G E O R G IA
P R IM A R Y A N D R U N O F F S E Q U E N C E S , W IT H B L A C K
C A N D ID A T E S
S U M M A R Y O F F IN D IN G S , E C O L O G IC A L R E G R E S S IO N
E C O L O G IC A L R E G . R E S U L T S
% B L A C K % W H IT E S Q U A R E D
V O T E R S V O T E R S C O R R E L A T IO N
V O T IN G V O T IN G C O E F F IC IE N T
F O R B L A C K F O R B L A C K (R2)
C A N D (S). C A N D (S).
C O U N T Y &
E L E C T IO N
BALD W IN COUNTY
1990 P R IM A R Y
G O V E R N O R 98 8 .95
1990 R U N O F F
G O V E R N O R 100 17 .95
BIBB COUNTY
1984 P R IM A R Y
B D . O F E D . 73 6 .98
1984 R U N O F F
B D . O F E D . 99 0 .93
1990 P R IM A R Y
G O V E R N O R 91 7 .98
1990 R U N O F F
G O V E R N O R 94 14 .97
DEKALB COUNTY
1976 P R IM A R Y
S E N A T E D. 43 85 28 .80
1976 R U N O F F
S E N A T E D. 43 100 22 .83
1978 P R IM A R Y
H O U S E D. 56 54 14 .80
80
APPENDIX 1
R A C IA L P O L A R IZ A T IO N IN G E O R G IA
P R IM A R Y A N D R U N O F F S E Q U E N C E S , W IT H B L A C K
C A N D ID A T E S
S U M M A R Y O F F IN D IN G S , E C O L O G IC A L R E G R E S S IO N
E C O L O G IC A L R E G . R E S U L T S
% B L A C K % W H IT E S Q U A R E D
V O T E R S V O T E R S C O R R E L A T IO N
V O T IN G V O T IN G C O E F F I C I E N T
F O R B L A C K F O R B L A C K (R 2)
C A N D (S ). C A N D (S).
C O U N T Y &
E L E C T IO N
1982 P R IM A R Y
H O U S E D. 56 41 14 .66
1984 P R IM A R Y
C O COMM. 69 13 .74
S E N A T E D. 43 65 24 .59
A P P E A L S CT. 59 51 .11
D E M . P R E S ID E N T 60 1 .96
1984 R U N O F F
C O COMM. 84 17 .78
S E N A T E D. 43 67 27 .65
1984 G E N E R A L
S E N A T E D. 43 99 50 .96
DEKALB COUNTY
1988 P R IM A R Y
H O U S E D. 51 67 34 .32
D E M . P R E S ID E N T 95 14 .89
1988 R U N O F F
H O U S E D. 51 80 17 .72
81
APPENDIX 1
R A C IA L P O L A R IZ A T IO N IN G E O R G IA
P R IM A R Y A N D R U N O F F S E Q U E N C E S ,
W IT H B L A C K C A N D ID A T E S
S U M M A R Y O F F IN D IN G S , E C O L O G IC A L R E G R E S S IO N
E C O L O G IC A L R E G . R E S U L T S
% B L A C K % W H IT E S Q U A R E D
V O T E R S V O T E R S C O R R E L A T IO N
V O T IN G V O T IN G C O E F F IC IE N T
F O R B L A C K F O R B L A C K (R2)
C A N D (S). C A N D (S).
C O U N T Y &
E L E C T IO N
G REENE COUNTY
1982 P R IM A R Y
H O U S E D IST . 106 61 0 .80
1982 R U N O F F
H O U S E D IST . 106 70 0 .86
1986 P R IM A R Y
H O U S E D IST . 106 89 0 .85
NEW TON COUNTY
1990 P R IM A R Y
G O V E R N O R 100 0 .98
1990 R U N O F F
G O V E R N O R 100 12 .95
82
APPENDIX 1
R A C IA L P O L A R IZ A T IO N IN G E O R G IA
P R IM A R Y A N D R U N O F F S E Q U E N C E S , W IT H B L A C K
C A N D ID A T E S
S U M M A R Y O F F IN D IN G S , E C O L O G IC A L R E G R E S S IO N
E C O L O G IC A L R E G . R E S U L T S
% B L A C K % W H IT E S Q U A R E D
V O T E R S V O T E R S C O R R E L A T IO N
V O T IN G V O T IN G C O E F F I C I E N T
F O R B L A C K F O R B L A C K (R 2)
C A N D (S ). C A N D (S).
C O U N T Y &
E L E C T IO N
PU TNAM COUNTY
1976 P R IM A R Y
S C H O O L B O A R D 43 20 .34
1976 R U N O F F
S C H O O L B O A R D 54 36 .39
1982 P R IM A R Y
H O U S E D. 106 52 0 .99
1982 R U N O F F
H O U S E D. 106 77 1 .99
1990 P R IM A R Y
G O V E R N O R 99 3 .99
1990 R U N O F F
G O V E R N O R 97 9 .99
TWIGGS COUNTY
1973 P R IM A R Y
O R D IN A R Y 83 0 .88
1973 R U N O F F
O R D IN A R Y 93 0 .92
1990 P R IM A R Y
G O V E R N O R 81 0 .95
1990 R U N O F F
G O V E R N O R 93 0 .94
83
APPENDIX 1
R A C IA L P O L A R IZ A T IO N IN G E O R G IA
P R IM A R Y A N D R U N O F F S E Q U E N C E S , W IT H B L A C K
C A N D ID A T E S
S U M M A R Y O F F IN D IN G S , E C O L O G IC A L R E G R E S S IO N
E C O L O G IC A L R E G . R E S U L T S
% B L A C K % W H IT E S Q U A R E D
V O T E R S V O T E R S C O R R E L A T IO N
V O T IN G V O T IN G C O E F F I C I E N T
F O R B L A C K F O R B L A C K (R2)
C A N D (S). C A N D (S).
C O U N T Y &
E L E C T IO N
W ARREN COUNTY
1984 P R IM A R Y
S C H O O L
S U P E R IN T E N D A N T 82 0 .99
1988 P R IM A R Y
SC H O O L
S U P E R IN T E N D A N T 78 8 .91
W ILKINSON COUNTY
1976 P R IM A R Y
BD. O F E D . 49 35 .02
1976 R U N O F F
B D . O F E D . 80 5 .82
1982 P R IM A R Y
CO COM M . 100 1 .86
1982 R U N O F F
CO COM M . 100 7 .89
1990 P R IM A R Y
G O V E R N O R 80 0 .94
1990 R U N O F F
G O V E R N O R 99 4 .96
84
F. TESTIMONY OF RONALD E. W EBER
[Excerpt from Trial Testimony, 5 T. Tr. p. 314, line
1-p. 326, line 19]
Q Is it true, Dr. Weber, that in attempting to evaluate
racial voting patterns and the opportunity for the elec
toral success of black preferred candidates in congres
sional democratic primary elections, that the most pro
bative elections to analyze are o ther congressional
democratic primary elections?
A If you have some. I guess—and I would add to that
if you’re—if you’re trying to evaluate the effect—the
potential effectiveness of African-American Districts,
you m ight w ant to s ta r t w ith elections in which an
African-American has competed against a non African-
American. You sta rt with that. I won’t say you’ll end
with that. You at least want to start with that.
Q Okay. And so, you might also want to look at other
democratic prim ary elections for sta te legislative or
executive offices because tha t would involve similar
dynamics? Is that correct?
A Yes, and that’s why I’ve looked at Andrew Young
versus Zell Miller. I’ve looked at Scott and Johnson in
that primary. I looked at the runoff there. Yes.
Q And would you consider Democratic presidential
preference primary elections to be somewhat less pro
bative but probably still relevant?
A I’d say they’re relevant. Yes.
Q But somewhat less so than the other elections that
I mentioned?
A If you’re going to have to weight things—in other
words, if you’re going to start assigning weights to the
85
average, you might discount the presidential primary a
little bit. Yes.
Q Well, don’t you assign weights to elections when
you’re deciding on elections to analyze in different cases
that you’ve testified in?
A Yes, but I think under the—under the—under the
—sort of the guidelines that your expert and I am oper
ating under, I th ink we would both agree th a t you
wouldn’t exclude that election. You would analyze it.
Q I’m not asking whether you would exclude it.
A Yeah. Yeah. Yeah. Yeah.
Q I’m asking about the weight being assigned to it.
A Yeah. What I’m saying is that if you pick a period
of which you’re going to analyze the elections, you
would choose all the elections.
Now, you might, as Dr. Lichtman does, you might say,
“Well, I ’m not going to analyze a certa in election
because the minority candidate didn’t get 15% of the
vote.” We didn’t have that here. These are all elections.
Q Right now we’re talking about types of elections
right now.
A Yes, sir. Yes, ma’am. Excuse me. I’m sorry.
Q I wore purple for that reason. The next category of
elections, if you had some non partisan elections for
legislative or executive offices, would you consider them
less helpful or less probative of voting patterns in con
gressional primary elections?
A You’d need to note them, and you’d need to know
whether or not those elections were being conducted at
the same time as the primary. And for example, the
86
three judicial races that we included in our report are
elections held at the same time as the primary.
So, the same elective that’s voting in the Democratic
primary is voting in those, but you also have interest
ingly the Republicans voting as well. So, you’re getting
a very conservative test that seems to me a polarization
and performance of a district.
Q Well, isn’t it true, Dr. Weber, that you’ve testified
to other federal courts that in fact it’s inappropriate to
take judicial elections into consideration when you’re
looking at congressional election voting patterns?
A It depends upon the state.
Q Well, I asked you isn’t that true that you’ve testi
fied to that in other cases?
A Yes, but it depends on the context in each state
whether you do that. In Texas, I did it. Everybody did it
in Texas. The indicator that the—that the state used as
one of its partisan indicators was based only on judicial
elections.
Q And in Texas, judicial elections are partisan; isn’t
that correct?
A That’s correct. Yes.
Q And-—but why did you say in Louisiana, Dr. Weber,
that judicial elections are totally irrelevant?
A I don’t believe I said they were totally irrelevant in
a congressional context. I said I didn’t do that. I did not
do the analysis. Or I had—the analysis was there. I did
not report the analysis. Basically, most judicial elections
in Louisiana are not hold at the same time as congres
sional elections. They’re for their—it’s a different elec
torate.
87
Q Isn’t it true that all elections in Louisiana are held
at non partisan primaries, at the same time?
A Yes, but they’re not held at the same time. Most
judges in Louisiana are elected in special elections held
a t tim es other than the regular elections. You have
hardly any judges elected at the regular elections.
Q In Louisiana, didn’t you testify, Dr. Weber, that you
didn’t consider judicial elections because voters have a
different calculus when they go to voting for a judge
than they do when voting in a congressional election?
A That’s part of it. Yes.
Q And would that be any different in Georgia?
A Would they have a different cal —
Q Do you have any reason to believe —
A — different calculus?
Q Do you have any evidence to show that calculus dif
ference that you testified to that somehow voters in
Georgia have the same calculus in voting for judges as
they do in voting in partisan congressional elections?
A Well—first of all, it’s not a partisan election you’re
making a comparison to. It’s a - - -
Q A partisan primary.
A It’s a partisan primary, which only Democrats —
Q Right.
A People who think of themselves as Democrats run.
Q Which is a different electorate than the judicial;
correct?
A Yes. But I’ve already admitted it. So you’ve got a
larger electorate participating in the judicial election.
88
Now the question is: Are the campaigns conducted
differently? Yes, they are. I’ll admit to that.
Q In other words, Dr. Weber, you applied a standard
pending on the particular circumstances of the case?
A No, I do not. I believe that the judicial elections in
Georgia provide a reasonably good indication of how
voters within the Eleventh Congressional District react
to African-American versus non-African-American can
didates.
Q In congressional elections?
A Not for congressional elections. You’re talking
about—we’re talking about the district—
Q W ell, isn’t that what you’re trying to analyze?
A We’re talking about hub of voters. No. We’re talk
ing about hub of voters.
Q So you’re not trying to analyze the opportunity of
black voters to elect a member of Congress in this dis
trict?
A You use, you use the information for that purpose,
yes.
Q Isn’t it true, Dr. Weber, that in Louisiana, the judi
cial elections th a t were a t issue generally did not
involve incumbent judges? That they involved white?
A Some did. Some did.
Q And that there were many, many elections that
involved black versus white challenges for an open judi
cial seat?
A Primarily, most of the elections in Louisiana are
about open seats; yes.
89
Q Okay. And isn’t it also true that in a Louisiana judi
cial election which you say are not relevant to analyzing
congressional elections there, you showed fairly high
levels of racial polarization if you did analyze the elec
tions?
A Say that again to me. Fm sorry.
Q Isn’t it true that in those Louisiana judicial elec
tions which the expert witness on the other side of the
case analyzed and you said were inappropriate, you had
fairly high levels of racial polarization?
A We had varying levels of racial polarization depend
ing on the part of the state.
Q Isn’t it true that in this case, Dr. Weber, the judicial
elections you analyzed all involved black incumbent
judges who had been appointed to their position?
A I believe they’ve all been appointed; yes.
Q And they were incumbents when they ran for elec
tion?
A They were able to put—if they did pictures in their
ads, they could put on their robes and say “re-elect the
judge.” Yes.
Q And they were all—each of those judges in fact has
the active support of the Governor of Georgia at the
time in their campaign?
A I believe they did; yes.
Q And isn’t it also true that the results of those judi
cial e lections showed su b s tan tia lly h igher w hite
crossover vote than the other partisan primary election
that you analyzed in this case?
A Substantially higher white crossover.
90
Q Higher. I will take the substantial out of it.
A Not always.
Q Can you give me an example where it didn’t?
A The primary in 1992 in District Eleven, a majority
of non A frican-A m ericans supported the A frican-
American candidates.
Q Well, let me ask you about that, Dr. Weber. You’re
talking about the initial primary.
HONORABLE JUDGE BOWEN: Ms. Murphy, I
want to compliment you on the depth of your explo
ration, but really isn’t the point made when you show us
that there is indeed a distinction between a judicial elec
tion and a congressional election.
MS. MURPHY: Well, Your Honor, I’m actually mov
ing on to a separate point now, if I may.
HONORABLE JUDGE BOWEN: Okay. Well, let
that same reasoning apply analogously to the next point.
MS. MURPHY: I’ll do my best.
Q (Mr. Murphy) You just mentioned the 1992 demo
cratic primary for Congress.
A Yes, that’s correct.
Q Weren’t there four African-American candidates in
that election?
A Yes, and one non African-American; yes.
Q And in fact in reporting your Plaintiffs’ “84 F ”, you
com bined the vo te to ta ls for those four A frican-
American candidates to look at white crossover vote;
did you not?
A Yes, that’s appropriate.
91
Q And isn’t it true, Dr. Weber, that you’ve opined
before that in fact it is improper, actually I can just read
you the—read you the testimony I’m talking about.
“In o rder to de te rm ine who is the candidate of
choice of black voters one needs to see which candidate
obtained the largest share of the black votes. There is
no other way to determine a group’s candidate of choice
and determine which candidate received more of their
votes than anyone e lse.. .”
And then you went on to say that you agree with Dr.
Lichtm an’s approach in th a t particu lar case, which
would put all the black candidates together for purposes
—you disagreed, I ’m sorry.
A Yes. Disagreed, yes.
Q With Dr. Lichtman’s approach?
A Yes. I think what you’re doing is you’re mixing
apples and o ranges here . In the original tab le in
Attachment E, I have done a separate estimation.
Q Correct.
A For the purposes of summarization, I have summa
rized that. The raw data is in the report. The reader can
easily see that and who got the first—who ran first, who
ran second; all those kinds of things.
The only criticism I made of Dr. Lichtman on that is
when, when you run the regression of all the candidates
combined, you can’t make those inferences. I mean,
that’s a relative minor criticism, it seems to me.
Q Well, but you’re saying here—you’re saying the
crossover vote for the AA candidate is 55.3%, But in
fact, no A A candidate in that election received a major
ity of the white vote; isn’t that true?
92
A I believe that’s true; yes. Yeah.
Q And I believe you testified in that same case that it
was not appropriate to lump all these black candidates
together in attem pting to determine the candidate of
choice of each racial group; isn’t that true?
A And that’s why I put candidates of choice, not can
didate of choice.
Q So what you’re saying is that when you want to
look at whether or not there’s racially polarized voting
in a Section 2 context, you can’t put the black candidates
together, but in this context is okay?
A No, I’m not saying that. All I’m saying is that it is
simply a function of how you summarize.
Q Exactly.
A It seemed to me it was ju st a simple function to
summarize this way than to present all of—
Q So you’re w illing to lump all th e cand idates
together for this purpose?
A Just for the purposes of this simple summarization.
Q Well, Dr. W eber, I ’d like for you to look a t
Plaintiffs’ Exhibit “84 F ”. Do you still have it in front of
you?
A No, I don’t.
HONORABLE JUDGE BOWEN: I think I have it.
Q I’d like for you to calculate for me, Dr. Weber the
average crossover vote if we don’t include the three
judicial elections and we don’t include the two elections
in which you in fact lumped th e black candidates
together to determine the white crossover vote?
Do you have a calculator?
93
A No. I don’t have a calculator to do it, because I’ve
already done for you.
Q Oh, you have?
A Yes. I knew you were going to ask this sort of
thing. All I’ve got to do is fine —
Q You should have just testified on direct. We could
have saved a lot of time.
A I’m trying to find it. I t’s underneath my supple
mental report. Give me just a moment to find it.
[Pause.]
A All right. Okay, here it is. Okay. Now, what’s your
question, please?
Q I’d like you to calculate the average crossover vote
when you exclude the three judicial elections and the
two elections where you in fact lump the black candi
dates together?
A Oh, I would have calculate that, because we’d now
be down to one —
Q Five.
A — two, three, four, five. Okay.
Q Well, first why don’t you tell me what it is when
you exclude the three judicial elections? I assume that’s
what you calculated.
A Yes. Yes. What I did is you take the judicial elec
tions out that crossover drops to 26.6%.
Q And what about if you take these two elections
where you lump the black candidates together?
A I don’t know why we’re doing this, but I’m going to
play along with you.
94
HONORABLE JUDGE BOWEN: Well, how long will
it take?
A I’ve got to add five numbers together and divide by
five.
HONORABLE JUDGE BOWEN: All right.
[Pause.]
A 17%.
Q Thank you. You mentioned a moment ago your sup
plemental report. In fact, your report has had three
incarnations in this case; hasn’t it? Your original report,
your revised report, and now we have a small supple
mental report that we received at the beginning of the
trial last week. Is that correct?
A Yes. I made some minor revisions before the depo
sition and added this.
Q Do you consider those revisions minor, Dr. Weber?
A The revisions that I gave you at the deposition?
Q Yes.
A Yes, I think they were minor.
Q Well, isn’t it true that you in fact did a substantially
d ifferen t analysis of the elections in the E leven th
Congressional District and adjusted for split precincts
in your revised report?
A Yes. Yes.
Q But you hadn’t done that in your original report?
A No, because I did not—when I got Dr. Lichtman’s
report and saw that we were getting substantially dif
ferent estimates of participation, I explored why there
were those differences. And I came to the conclusion
95
that split — treatm ent of split precincts made a differ
ence in participation estimates.
Q And you determined to do an adjustment for those
split precincts after reading Dr. Lichtman’s report; is
that correct?
A Yes.
Q And you determ ined th a t ad ju stin g for sp lit
precincts in fact gave you more accurate results?
A For participation; not necessarily for polarization.
Q But the results you rely on, at least in comparing
your summary exhibits are adjusted for split precincts;
isn’t that correct?
A I believe they are; yes.
Q So you consider those more reliable results in your
report?
A Yes.
96
G. R E V IS E D R E P O R T O F R O N A L D E . W E B E R
[Plaintiffs’ Exh. 82, at cover-4,22, f 39-32, If 49
and Attach. E]
W IL D E R C R A N E P R O F E S S O R O F G O V E R N M E N T
U N IV E R S IT Y O F W IS C O N S IN — M IL W A U K E E
J u n e 16,1994
INTRODUCTION
A. Background and Experience
1. My name is Ronald E. Weber and I am a resident
of the State of Wisconsin.
2. I am curren tly the W ilder Crane Professor of
G overnm ent and C hairm an of the D ep artm en t of
P o litical Science a t the U n iv e rs ity of W isconsin,
Milwaukee, Wisconsin; P residen t of Campaign and
Opinion Research Analysts, Inc.; and former co-editor of
The Journal o f Politics. I received my B.A. in Political
Science and History from Macalester College, St. Paul,
MN, in 1964 and a Ph.D. in Political Science from
Syracuse U n iv e rs ity in 1969, w ith sp ec ia lties in
American state politics, voting behavior, and quantita
tive analyses of political data. A copy of my curriculum
vitae is attached as Attachment A.
3. I am the author of numerous scholarly works on
sta te political behavior, including several works on
state legislative elections and voting behavior at the
individual and aggregate levels of analysis. These
works have appeared in such academic journals as the
Am erican Political Science Review, The Journal o f
Politics, Midwest Journal o f Political Science, Public
Opinion Quarterly, and Legislative Studies Quarterly.
4. I have been retained as a consultant and expert
witness in a number of voting rights cases and have
been qualified as an expert by the U.S. District Courts
97
in the M iddle D is tr ic t (N o rth ern and S ou thern
Divisions) of Alabama, the N orthern D istrict (Tal
lahassee Division) and Middle District of Florida, the
Eastern, Middle, and W estern Districts of Louisiana,
the District of Maryland, the District (Western Division)
of M assachusetts, the E as te rn D istric t (Southern
Division) of Michigan, the Northern District (Eastern
and Western Divisions) of Mississippi, and the Northern
District of Texas. I have given testimony by deposition
in a num ber of cases, including a deposition for the
plaintiffs in the Congressional redistricting case of Shaw
v. Reno, n/k/a Shaw v. Hunt (Eastern District of North
Carolina, Raleigh Division). A listing of the cases in
which I have testified in Federal court or I was deposed
under oath is attached as Attachment B. I also have
extensive experience developing redistricting plans for
local and state government clients and assisting them
with preclearance of those plans under Section 5 of the
U.S. Voting Rights Act of 1965, as amended in 1982.
5. I have been retained by plaintiffs in this case and
am being compensated at the rate of $100 per hour plus
out-of-pocket expenses.
B. Purpose o f the Analysis
6. I have addressed the following questions in analyz
ing w hether the 1992 Congressional red istricting in
Georgia for the Eleventh Congressional district results
in a v iolation of the F o u rte en th and F if te e n th
Amendments of the U.S. Constitution in accord with
factors set forth by the U.S. Supreme Court in Shaw v.
Reno:
(1) whether or not race was the overriding factor used
by the State of Georgia to draw the boundaries of the
Eleventh Congressional district;
98
(2) whether or not the Eleventh Congressional dis
trict is compact;
(3) whether or not the Eleventh Congressional dis
trict violates other traditional districting criteria;
(4) w h e th e r or not voting p a tte rn s in recen t
Democratic primary, runoff, and general elections
held within the boundaries of the E leventh Con
gressional district reveal patterns of participation
differences betw een African-Americans and non-
African Americans and racially polarized voting,
and, if so, to what extent;
(5) w h e th er or no t th e E lev en th C ongressional
District is overly safe from the standpoint of assuring
the election of a candidate of choice of African-
American voters;
(6) whether or not more narrowly tailored plans were
or can be created for an Eleventh Congressional dis
trict that can achieve the same electoral results as the
state plan that are less violative of traditional district
ing criteria; and
(7) whether or not the state plan for the Eleventh
Congressional district results in a maximization of
A frican-A m erican rep resen ta tio n in the Georgia
congressional delegation.
7. In attem pting to assist the court in addressing
these questions, I have undertaken an analysis of data
reported in the State of Georgia’s 1991-92 preclearance
submissions to the U.S. Department of Justice, as well
as other demographic and electoral data. The results of
my analysis to date will be presented in this report in
the following form: in section I, I will describe the analy
ses conducted to answer the first three questions and
set forth my conclusions on those questions; in Section
99
II, I will describe the electoral database employed, my
analytical methodology for the participation and polar
ization analyses, and set forth my conclusions based
upon the application of this methodology to this data
base; and in section III, I will describe the analyses con
ducted to answer the last two questions and set forth
my conclusions on those questions. Tables and exhibits
relevant to my analyses will be included as attachments
to this report.
2. Racial Polarization in Voting
39. I also analyzed the eleven contested general elec
tions for U.S. Congress in 1992 in Georgia to determine
the patterns of racially polarized voting within each dis
trict (see Attachment E). In the seven district elections
for which valid African-American voter behavior esti
mates can be obtained, the African-American voters dis
play cohesive support for the Democratic nominee in
each district. In the Eleventh Congressional district, the
African-American voters give an estimated 96 percent
of th e ir support to the Democratic p a rty nominee,
Cynthia McKinney. In the other two majority African-
American districts, the African-American voters give
between 93 and 100 percent of their support to African-
American Democratic nominees in the general election.
In the eight non-African-American majority districts,
the non-African-American voters give a m ajority of
their support to Republican nominees in six of the eight
districts. In only the Seventh and Ninth districts do
non-African-Americans give a majority of their support
to Democratic nominees. Non-African-American voters
in the three African-American majority districts act
similarly to non-African-American voters in other dis
tricts, giving a majority of their support to Republican
100
nominees. The non-African-American cross-over vote in
the African-American majority districts ranges from
about 23 percent in the Second district to 44 percent in
the Fifth district. The estimated figure for the Eleventh
Congressional district is 28.6 percent.
40. I also analyzed the D em ocratic p rim ary and
runoff p rim ary resu lts for Congress in 1992 in the
precincts contained in the th ree African-Am erican
majority Congressional districts of Georgia to de ter
mine the p a tte rn s of racial polarization in voting in
those elections. In the D em ocratic p rim ary in the
Eleventh Congressional district, Cynthia McKinney was
the plurality choice of African-American voters with an
estimated 40-41 percent level of support, while George
Deloach was th e p lu ra lity choice of non-A frican-
American voters with an estimated level of support of
44-45 percent. The combined vote for the four African-
American candidates was an estimated 55-56 percent of
the vote cast by non-African-Americans, indicating a
low level of racial polarization in voting in that election.
In the Democratic runoff primary, African-Americans
were cohesive in their support of Cynthia McKinney,
giving her an estimated 90-96 percent level of support.
H er cross-over support from non-African-Americans
was estimated at 23-24 percent. With the combination of
African-American and non-African-American support,
she captured over 55 percent of the vote. Therefore, no
legally or politically consequential racial polarization in
voting occurred in that election.
41. In the Second Congressional district, African-
Americans gave plurality support to Sanford Bishop at
an estimated level of 31.7 percent in the Democratic pri
mary. Charles Hatcher, the incumbent Congressman,
was the choice of non-African-Americans, a t an esti
mated level of support of 66.4 percent. In the runoff
101
primary, Bishop was the choice of African-Americans
receiving an estimated level of support of 70.5 percent.
His cross-over support from non-African-Americans
was an estimated 22.3 percent. Bishop was the victor in
the runoff primary, thus resulting in no legally or politi
cally consequential racial polarized voting. In the
Democratic primary for the Fifth Congressional district,
John Lewis, the incumbent, was the choice of both
African-American and non-African-American voters.
42. I also analyzed the Democratic primary runoff
results in 1990 for Governor and the Democratic pri
mary results in 1988 for President in the precincts con
ta in ed in th e th re e A frican-A m erican m ajo rity
Congressional districts of Georgia to determine the pat
terns of racial polarization in voting in those elections.
A frican-A m erican voters give cohesive support to
Andrew Young, the former Congressman and Mayor of
Atlanta, in the Democratic runoff for Governor in 1990
in the Eleventh, Second, and Fifth Congressional dis
tricts, with the African-American crossover vote run
ning as high as an estimated 18.3 percent in the Fifth
Congressional district. Cross-over voting from non-
African-Americans in the three Congressional districts
ranges from a high of about 35 percent in the Fifth dis
trict to lows of eight to eleven percent in the Second and
Eleventh districts respectively. While the 1990 Demo
cratic runoff election results indicate some racial polar
ization in voting, the cross-over voting by each group
negates legally or politically consequential racial polar
ized voting in each district.
43. African-American voters also gave cohesive sup
port to Jesse Jackson in the Democratic Presidential
Preference primary in 1988 with the African-American
crossover vote running about an estimated six to seven
p e rcen t in the th ree A frican-A m erican m ajo rity
102
Congressional districts. Cross-over voting from non-
African-Americans in the three Congressional districts
was minimal, with the election results indicating some
racial polarization in voting. The net effect of participa
tion patterns of the two groups in the election, with the
African-American voters participating at a higher rate
than non-African-Americans, advantaged the African-
American voters thus making the presence of racial
polarization in voting of no legal or political conse
quence.
44. I have also analyzed the patterns of racial polar
ization in voting in the 1992 Democratic primary and
runoff elections for the state Commissioner of Labor
within the precincts contained in the th ree African-
A m erican m ajority Congressional d is tric ts . In the
E leventh Congressional district, African-Americans
were cohesive in support of African-American candidate
Scott in the Democratic primary election. Non-African-
Americans gave a slight majority of their support to
Poythress the non-African-American candidate in the
election. In the runoff primary, Scott received cohesive
support from A frican-A m erican v o ters while non-
African-American cross-over support for Scott was
estimated at about 27 percent. Although there is evi
dence of some racial polarization in voting in this con
test, it is not legally or politically consequential because
the candidate preferred by African-American voters
wins within the boundaries of the Eleventh Congres
sional district. N either African-Americans nor non-
African-Americans display cohesive patterns in racial
polarization in voting in the Democratic primary for
s ta te C om m issioner of L abor in th e Second
Congressional district. Scott is preferred by a majority
of African-American voters, while Poythress is sup
ported by a plurality of non-African-Americans within
the district. Voting is somewhat racially polarized in the
103
runoff election in the Second Congressional district,
w ith P o y th re ss w inning th e e lection w ith in the
precincts in the district. No legally or politically conse
quential racial polarization in voting occurs in the state
C om m issioner of L abor elections held w ithin the
precincts of the Fifth Congressional district in that in
both the Democratic primary and runoff elections, Scott
who is supported by African-American voters would
have been the winner of the election.
45. I have also analyzed the patterns of racial polar
ization in voting in the three recent statewide non-parti
san elections for the state judiciary within the precincts
contained in the th ree A frican-A m erican m ajority
Congressional districts. These are three elections in
which African-American candidates were elected to full
terms on Georgia’s Supreme Court or Court of Appeals.
These non-partisan elections were held at the time of
the primary elections in July. In the Eleventh Congres
sional district, African-Americans were cohesive in sup
port of the African-American candidates in all three
elections. Cross-over support by non-African-Americans
played a major role in all three elections, with Robert
Benham receiving a majority of non-African-American
support in the 1990 election for Supreme Court justice.
N on-A frican-A m erican suppo rt for th e o ther two
African-American candidates was strong also with Leah
Sears-Collins receiving an estimated 42.3 percent in the
1992 election and Clarence Cooper obtaining an esti
mated 44.5 percent support in the 1990 election. Racial
po lariza tion in voting does not occur in the 1990
Supreme Court election since both groups favor the
same candidate. And the degree of racial polarization in
voting is low in the other two judicial contests. Legally
or politically consequential racial polarization in voting
does not occur in th a t the candidates preferred by
104
African-American voters win all three elections held
within the boundaries of the Eleventh Congressional
district. A similar pattern in racial polarization in voting
also occurs in the Second Congressional district. The
1990 Supreme Court election is not polarized, while the
other two elections display low levels of polarization.
Again the polarization in the two elections has no legal
or political consequence as the candidates of choice of
African-Americans prevail. No racial polarization in vot
ing occurs in the three judicial elections held within the
precincts of the Fifth Congressional district in that the
two groups support the same candidate of choice.
46. I have also analyzed the pattern of voting in sev
eral 1986,1990, and 1992 Democratic primary and runoff
elections for State Senate and House seats held within
the precincts contained in the Eleventh Congressional
districts to determine the extent of racial polarization in
voting in sub-regions of the district. In the most recent
Democratic primary elections of 1992, voting was some
w hat polarized for the 22nd S ta te Senate D istric t
(Burke and Richmond Counties), the 66th State House
D is tric t (DeKalb County), the 111th S ta te House
D is tr ic t (G reen, Pu tnam , T alia ferro , and W ilkes
C ounties), and th e 116th S ta te H ouse D is tric t
(Richmond County). Voting was not polarized in the
D em ocratic p rim ary for S ta te Senate D is tr ic t 55
(DeKalb County) because the candidate preferred by
both groups was victorious. The racially polarized vot
ing was legally or politically consequential in only one
district election in 1992—the runoff primary in House
District 66 (DeKalb County) where the candidate pre
ferred by a m ajority of non-African-Americans won.
(However, the African-American participation in that
runoff primary was so low even cohesion behavior by
those African-Americans voting was insufficient to elect
a candidate of choice.) A lack of African-American cohe-
105
sion in the 111th House District resulted in the defeat of
the candidate preferred by African-American voters in
that year.
47. My analyses of the two elections from 1990 indi
cate a pattern of some racial polarization in voting in
both elections. The candidate of choice of African-
Am ericans was elected in S ta te House D istric t 56
(DeKalb County) despite racial polarization in voting.
Thus, the racial polarization in voting was legally or
politically consequential only in State Senate District 22
(Richm ond C ounty). H ow ever, th a t S ta te Senate
District was not majority African-American in total or
voting age population at that time and the patterns of
participation and voting by the two groups were insuffi
cient to produce a win for the candidate of choice of
African-American voters. The analyses of the three
State House elections from 1986 indicate that voting
was somewhat racially polarized in Districts 106 and 116
and not polarized in District 56. The pattern of racial
polarization in voting was not legally or politically con
sequential in Districts 106 and 116 since the preferred
candidate of African-Americans won both elections.
48. The analyses of racial polarization in voting in the
different recent elections held within the boundaries of
the Eleventh Congressional district indicate that some
racial polarization in voting occurs, but for the most part
the racial polarization in voting is not legally or politi
cally consequential. Candidates of choice of African-
American voters won all contests for the Congressional
office in 1992, won the Democratic runoff for Governor
in 1990, won the Democratic Presidential Preference
prim ary in 1988, and won the Democratic runoff for
state Commissioner of Labor in 1992, and won the three
positions on the state judiciary in 1990 and 1992. In the
selected elections for S tate Senate and S tate House
106
positions, candidates preferred by African-Americans
lost only two times when the African-American group
was cohesive in the election—Senate District 22 in 1990
and House District 66 in 1992. In the remaining eight
other elections for State Senate or House positions, the
level of racially polarized voting was not legally or polit
ically consequential.
3. Reconstituted Election Return Analysis
49. Using actual precinct-level returns for the 1988,
1990, and 1992 elections analyzed above, it is possible to
reconstitu te the resu lts of s ta te elections into the
boundaries of the Congressional districts adopted by
Georgia in 1992. These reconstituted results for the 1990
Democratic runoff primary for Governor and the 1988
Democratic Presidential Preference primary indicate
th a t th e A frican-A m erican p re fe rre d cand idates
Andrew Young (1990) and Jesse Jackson (1988) would
have been the winners of those elections in the Eleventh
Congressional district as well as the Second and Fifth
Congressional districts if the elections were held only
within the boundaries of the 1992 Congressional dis
tricts (see Attachment F). In addition, I have reconsti
tuted the results of the 1990 and 1992 state judicial con
tes ts as well as the two 1992 sta te Commissioner of
Labor elections to see how African-American preferred
candidates fared w ithin the precincts of the th ree
Congressional districts. The results of the three state
judicial contests reconstituted to the precincts of the
three African-American m ajority Congressional dis
tric ts indicate tha t all three African-American candi
dates win the three Congressional districts, with the
w inning p e rcen t being qu ite com fortable in the
Eleventh Congressional district. In the reconstituted
re su lts for th e s ta te C om m issioner of L abor
107
Democratic primary election, Scott wins a majority of
the vote in the Democratic primary in the Eleventh and
Fifth Congressional districts while winning a plurality
of the vote in the Second Congressional district. In the
runoff election, Scott is elected in the E leventh and
Fifth Congressional districts, while losing in the Second
Congressional district. These analyses reveal that the
candidates p re fe rred by A frican-A m ericans often
achieved very comfortable margins of victory in each of
the contests held within the precincts in the Eleventh
Congressional district. The patterns of group participa
tion, African-Am erican cohesion, and non-African-
American cross-over voting all contributed to these
margins of victory.
STATE OF GEORGIA
1992 Democratic Run-Off
11th Congressional District
(% Of REG)
Black Nonblack Corr.
Weighted Candidate Vote Vote Coef R2
D eloach .021(9.3% ) .205(77.1% ) -.440** .19
M cK inney (B) + .204 (90.7%) .061 (22.9%) .496** .25
.225 .266
Extreme Cases Candidate Black Vote Nonblack Vote
D eloach .025 (12.0%) .134 (76.6%)
M cK inney (B) + .184 (88.0%) .041 (23.4%)
.209 .175
108
Un- Black Nonblack Corr.
Weighted Candidate Vote Vote Coef. R2
D eloach .023(10.5% ) .233(79.0% ) -.466** .22
M cK inney (B) + .196(89.5% ) .062(21.0% ) .455** .21
,219~ .295
(B) d e n o te s b lack c a n d id a te
+ d e n o te s w in n e r
** S ta tis tic a lly sign ifican t a t th e .01 level
STATE OF GEORGIA
1992 Democratic Run-Off, 11th Congressional District
Adjusted for Split Precincts
(% Of REG)
Black Nonblack Corr.
Weighted Candidate Vote Vote Coef. R2
D eloach .009(4.3% ) .247(76.0% ) -.543** .29
M cK inney (B) + .199 (95.7%) .078 (24.0%) .419** .18
.208 .325
Extreme Cases Candidate Black Vote Nonblack Vote
D eloach .026 (12.1%) .197 (76.7%)
M cK inney (B) + .188(87.9% ) .060(23.3% )
.214 .257
109
Un
weighted Candidate
Black Nonblack Corr.
Vote Vote Coef R2
D eloach .022(10.1% ) .250(77.9% ) -.508** .26
M cK inney (B )+ .195(89.9% ) .071(22.1% ) .426** .18
.217 .321
(B) d e n o te s b lack can d id a te
+ d e n o te s w in n e r
** S ta tis tic a lly sign ifican t a t th e .01 leve l
STATE OF GEORGIA
1992 11th Congressional District
1990 Democratic Run-Off for Governor
(% Of REG)
Weighted Candidate
Black Nonblack Corr.
Vote Vote Coef. R2
M iller
Y o u n g (B )+
.049(10.4% ) .294(89.1% ) -.576** .33
.420(89.6% ) .036(10.9% ) .737** .54
.469 .330
Extreme Cases Candidate Black Vote Nonblack Vote
.054 (11.3%) .287 (81.1%)
.426 (88.7%) .067 (18.9%)
M iller
Y oung (B) +
.480 .354
110
Un- Black Nonblack Corr.
Weighted Candidate Vote Vote Coef. R2
M iller .098(16.9% ) .307(99.4% ) -.291** .08
Y oung ( B ) + .483(83.1% ) .002(0 .6% ) .483** .23
.581 .309
(B) d e n o te s b lack c a n d id a te
+ d e n o te s w in n e r
** S ta tis tic a lly sign ifican t a t th e .01 level
STATE OF GEORGIA
1992 11th Congressional District
1988 Democratic Presidential Preference Primary
(% Of REG)
Black Nonblack Corr.
Weighted Candidate Vote Vote Coef. R2
W h ite
D e m o c ra ts .024 (5.9%) .151 (105.6% ) -.779** .61
Jac k so n (B) + .380 (94.1% ) -.008 (-5.6% ) .905** .82
.404 .330
Extreme Cases Candidate Black Vote Nonblack Vote
W h ite
D e m o c ra ts .025 (6.1%) .148 (80.9%)
Ja c k so n (B) + .386 (93.9%) .035 (19.1%)
.411 .183
I l l
Un- Black Nonblack Corr.
Weighted Candidate Vote Vote Coef. R 2
W h ite
D e m o c ra ts .026(6.5% ) .157 (105.4% )-.736** .54
Ja c k so n (B) + J373 (93.5%) -M 8 (-.5.4%) .873** .76
.399 .149
(B) d e n o te s b lack c a n d id a te
+ d e n o te s w in n e r
** S ta tis tic a lly s ig n ifican t a t th e .01 lev e l
STATE OF GEORGIA
1992 Democratic Run-Off for Commissioner of Labor
11th Congressional District
(% Of REG)
Black Nonblack Corr.
Weighted Candidate Vote Vote Coef. R 2
P o y th re s s .018 (8.3%) .223 (72.9%) -.552** .30
S c o tt ( B ) + .199(91.7% ) .083(27.1% ) .401** .16
~211 .306
Extreme Cases Candidate Black Vote Nonblack Vote
P o y th re s s .023 (10.6%) .182 (73.1%)
S c o tt (B) + .194 (89.4%) .067 (26.9%)
.217 .249
112
Un- Black Nonblack Corr.
Weighted Candidate Vote Vote Coef. R2
P o y th re s s .019(8.7% ) .235(73.7% ) -.543** .30
S c o tt ( B ) + .199(91.3% ) .084(26.3% ) .378** .14
.218~ .319~
(B) d e n o te s b lack can d id a te
+ d e n o te s w in n e r
** S ta tis tic a lly sign ifican t a t th e .01 level
STATE OF GEORGIA
1992 Non-Partisan Election for Supreme Court Justice
11th Congressional District
(% Of REG)
Weighted Candidate
Black
Vote
Nonblack
Vote
Corr.
Coef. R2
B osw ell .029 (12.4%) .135 (57.7%) -.492** .24
S ears-
C ollins (B) + .205 (87.6%) .099 (42.3%) .344** .12
.234 .234
Extreme Cases Candidate Black Vote Nonblack Vote
B osw ell .031 (15.3%) .132 (55.5%)
S ears-
C ollins (B) + .171 (84.7%) .106 (44.5%)
.202 .238
113
Un
weighted Candidate
Black
Vote
Nonblack
Vote
Corr.
Coef. R2
B osw ell
S ea rs-
.023 (11.8%) .162 (55.9%) -.296** .09
C ollins (B) + .172 (88.2%)
U 5
.128 (44.1%)
.290
.089 .01
(B) d e n o te s b lack can d id a te
+ d e n o te s w in n e r
** S ta tis tic a lly sig n ifican t a t th e .01 level
STATE OF GEORGIA
1990 Non-Partisan Election for Supreme Court Justice
1992 11th Congressional District
(% Of REG)
Black Nonblack Corr.
Weighted Candidate Vote Vote Coef. R2
B en h am (B) + .155(74.5% ) .133(59.4% ) .077** .01
S a lte r .053 (25.5%) .091 (40.6%) -.247** .06
.208 .224
Extreme Cases Candidate Black Vote Nonblack Vote
B en h am (B) + .130 (75.1%) .124 (58.8%)
S a lte r .043 (24.9%) .087 (41.2%)
.173 .211
114
Un- Black Nonblack Corr.
Weighted Candidate Vote Vote Coef. R2
B en h am (B) + .136(75.6% ) .126(58.3% ) .035 .00
S a lte r .044(24.4% ) .090(41.7% ) -.276** .08
.180 .216
(B) d e n o te s b lack c a n d id a te
+ d e n o te s w in n e r
** S ta tis tic a lly sign ifican t a t th e .01 level
STATE OF GEORGIA
1990 Non-Partisan Election for Court of Appeals
1992 11th Congressional District
(% Of REG)
Black Nonblack Corr.
Weighted Candidate Vote Vote Coef. R2
B lack b u rn .041(18.9% ) .117(55.5% ) -.501** .25
C o o p er ( B ) + .176(81.1% ) .094(44.5% ) .250** .06
.217 .211
Extreme Cases Candidate Black Vote Nonblack Vote
B lack b u rn .031 (18.1%) .100 (52.6%)
C o o p er (B ) + .140 (81.9%) .090 (47.4%)
.171 .190
115
Un- Black Nonblack Corr.
Weighted Candidate Vote Vote Coef. R®
B lack b u rn .039(21.5% ) .116(55.2% ) -.410** .17
C o o p er ( B ) + .142(78.5% ) .094(44.8% ) .159* .08
.181 " H o
(B) d e n o te s b lack c a n d id a te
+ d e n o te s w in n e r
* S ta tis tic a lly sign ifican t a t th e .05 leve l
** S ta tis tic a lly sig n ifican t a t th e .01 leve l
116
H. JOINT STATEM ENT OF UNDISPUTED FACT
3. The 1990 Decennial Census reports that the State
of Georgia’s population is 6,478,216 persons of whom
1,746,565 (26.96%) are black persons. Abrams Exhibit 1.
4. The to ta l voting age population of Georgia is
4,750,913, of which 3,490,414 (73.5%) are white and
1,168,142 (24.6%) are black. 1990 Census, Table 1, at p. 1.
Abrams Exhibit 1.
61. The m em bers of th e G eorgia House of
Representatives are elected from 180 single-member
districts. Of those districts 42 have a majority black
total population and 138 have a majority white total
population. Thirty of the 42 (71.4%) majority-black dis
tricts are represented by black members and 12 of the
42 (28.6%) are represented by white members. Of the
138 majority-white House districts, one (0.7%) is repre
sented by a black member. Exhibits U.S. 57, Abrams 24.
62. Thirty of the thirty-one (96.8%) black members of
the House are elected from majority-black districts.
Exhibits U.S. 57, Abrams 24.
63. The members of the Georgia Senate are elected
from 56 single-member districts. Of those districts 13
have a majority black total population and 43 have a
majority white total population. Nine of the thirteen
(69.2%) m ajority-black d istric ts are represen ted by
black members and four of the 13 (30.8%) are repre
sented by white members. Of the 43 majority-white
Senate districts, none are represented by a black mem
ber. Exhibits U.S. 57, Abrams 23.
64. One hundred percent of the black members of the
Georgia Senate are elected from majority-black dis
tricts. Exhibits U.S. 57, Abrams 23.
117
VOTING H ISTO RY
76. In 1908, the Georgia General Assembly passed a
proposed amendment to the Georgia Constitution to
prescribe the qualifications for electors and provide for
the registration of voters. The proposed amendment
established requirements for eligibility to vote, includ
ing state and county residence requirements, payment
of taxes due, and satisfaction of one of the following:
honorable service in one of several specified wars, a
descendent of the former, good character and under
standing of the duties and obligations of citizenship
under a republican form of government, ability to read
and w rite any p a rag rap h of th e U nited S ta te s or
Georgia Constitution (with a proviso for those unable to
do so because of physical disability), or ownership of
property of a specified character. In the published, writ
ten version of the June 24, 1908 message of Governor
Hoke Smith to the Georgia General Assembly, the
remarks concerning the aforesaid proposed amendment
are reported under the heading “Disfranchisement Act.”
No such amendment is in force in Georgia.
77. The poll tax was a feature of Georgia electoral law
in the years immediately prior to 1945, but it was not in
1945 and thereafter.
78. Between 1900 and 1944, the democratic primary
was a part of the political process that existed in the
state, and blacks were excluded therefrom. Exclusion
from the democratic primary did not, however, exclude
black persons from voting in any other elections, includ
ing general elections, non-partisan candidate elections,
special elections, referenda, bond elections, republican
primaries and the like.
79. The district court in King v. Chapman, 62 F. Supp.
639 (M.D. Ga. 1945), affd , 154 F.2d 460 (5th Cir.), cert.
118
denied, 327 U.S. 800 (1946), found unconstitutional the
exclusion of black citizens from participation in the
D em ocratic p rim ary election in Muscogee County,
G eorgia, which exclusion had been p u rsu a n t to a
Democratic party rule of state-wide applicability. Said
rule has not been in force since 1946.
80. The district court in King v. Chapman found that
“[t]he Democratic Party is the dominant and controlling
political party in Georgia. No other party has held a
statewide primary during the past 40 years. Since 1900
Democratic nominees for United States Senator, mem
bers of the House of Representatives, Governor and
other Statehouse officers, nominated at primaries, have
been elected in the ensuing general election.” During
the time referred to in the quotation, a candidate win
ning in the Democratic primary for the offices listed in
the quotation usually went on to wrin in the general elec
tion, although there was no Democratic prim ary for
many elections, such as general elections, special elec
tions, non-partisan elections and various referenda.
81. Generally speaking, more blacks were registered
and voted in the State of Georgia in the years after the
King v. Chapman decision than before. The number of
black voters has generally increased in the state over
the years.
82. In 1949, following the abolition of the w hite
Democratic primary in 1945 and the election of Herman
Talmadge as Governor in November 1948, the Georgia
General Assembly enacted a “Voters’ Registration Act.”
The 1949 Voters’ Registration Act, among other things,
declared existing registration lists null and void (except
for special elections held prior to the next general elec
tion), req u ired the re -re g is tra tio n of vo te rs , and
required further that citizens vote at least once in every
119
two years to maintain voter eligibility. This Act was
repealed by 1958.
83. Under the 1949 Voter’s Registration Act, voters
could qualify to register to vote by demonstrating an
ability to read and write (this requirement could be sat
isfied if the applicant was unable to read solely because
of physical disability), or by reason of good character
and understanding of the duties and obligations of citi
zenship under a Republican form of governm ent by
answering 10 of 30 questions. These provisions are no
longer in force in Georgia.
84. The Georgia General Assembly in 1950 enacted
legislation, which, among other things, permitted voters
who had registered under prior provisions of Georgia
law, i.e., those registered prior to 1949, to remain eligi
ble to vote under certain circumstances. These provi
sions are no longer in force in Georgia.
85. In 1958, the Georgia General Assembly passed a
“Voters Registration Act” which readopted some of the
provisions of the 1949 Voters Registration Act, and,
among other things, required for those illiterate persons
seeking to register that 20 of 30 questions be answered
correctly. These provisions are no longer in force in
Georgia.
86. For some years prior to April 28,1962, the State of
Georgia operated under a county unit system for deter
mining the nominees in primary elections for United
States Senator and state officers elected on a state-wide
basis, including justices of the Georgia Supreme Court
and judges of the Georgia Court of Appeals.
87. Under the county unit system these nominations
were not determ ined on the basis of popular vote.
Rather, prior to 1962, each county was assigned a cer-
120
tain number of “units” based on the number of represen
tatives elected to the lower House of the Georgia legis
lature from the county, and a candidate for one of the
offices listed in the previous request for admission won
a county’s unit votes by receiving a plurality of the pop
ular votes cast in that county for that office.
88. The effect of the county unit system was generally
to attribute certain representation and political influ
ence to counties, in part regardless of their population,
and some part of the consequence of that system was to
give equal or proportionately greater representation on
a per capita basis to the less populated, rural counties.
89. As late as 1962, Democratic Party rules in Georgia
allowed the Democratic Executive Committee of each
congressional district to decide whether to employ the
county unit system or a popular election plan for the pri
mary nomination of candidates for Congress. Rules and
R eg u la tio n s o f the S ta te D em ocratic E xecu tive
Committee o f Georgia (Adopted April 18, 1962), at 20-
21. An authentic copy of this document is Exhibit U.S.
53.
90. According to historian James C. Bonner, whose
affidavit was filed on behalf of the plaintiff in Sanders v.
Gray, 203 F. Supp. 158 (N.D. Ga. 1962), supporters of
the county unit system characterized it as a barrier
against the “bloc vote” and “race mixing,” and as a bul
wark for defense of “the Southern way of life.” James C.
Bonner, “Legislative Apportionment and County Unit
Voting in Georgia Since 1777,” Georgia H istorical
Ouarterly, XLVII, 351-74, at 370 (December 1963).
91. According to historian Numan V. Bartley, support
for the county unit system was correlated with support
for racial discrimination. “Conservatives consistently
supported the protection and promotion of the county
121
unit system in politics and white supremacy in the social
order.” Numan V. Bartley, From Thurmond to Wallace:
P o litica l Tendencies in Georgia. 194.8-1968, 22
(Baltimore, Md., The Johns Hopkins University Press,
1970).
92. A ccording to h is to rian N um an V. B artley ,
Congressman James C. Davis of the Fifth district was a
well-known advocate of racial discrimination: “Davis
possessed a reactionary voting record in the House of
R epresen ta tives and was a fiery proponent of the
virtues of white supremacy.” Numan V. Bartley, From
Thurmond to Wallace: Political Tendencies in Georgia.
1948-1968, 52 (Baltim ore, Md., The Johns Hopkins
University Press, 1970).
93. H isto rian N um an V. B artley has described
C harles W eltner’s politics in the following term s:
“W eltner espoused a generally progressive program
th a t con trasted sharp ly w ith D avis’ congressional
record and campaign oratory.” Numan V. Bartley, From
Thurmond to Wallace: Political Tendencies in Georgia.
1948-1968, 52 (Baltim ore, Md., The Johns Hopkins
University Press, 1970).
94. In 1964, the state of Georgia was required by the
U.S. Supreme Court to reapportion its congressional
districts in accordance with the one person, one vote
principle. Wesberry v. Sanders, 376 U.S. 1 (1964).
95. In A pril, 1962, a th ree -ju d g e d is tr ic t court
enjoined the State of Georgia’s use of the county unit
system, as it was then implemented. Sanders v. Gray,
203 F. Supp. at 171. On appeal, the Supreme Court
struck down the county unit system in Georgia as viola
tive of the one-person, one-vote rule. Gray v. Sanders,
372 U.S. at 376-81 (1963).
122
96. On June 24, 1964, the State of Georgia enacted a
voting code which established, comprehensive voting
procedures, mechanisms and requirem ents applicable
uniformly throughout the state. The law required candi
dates for nomination or election to federal, state and
county offices in Georgia to obtain a m ajority of the
votes cast to win the nomination for public office. If no
candidate receives a majority of the vote, the candidates
with the two highest number of votes run again in a run
off primary or election. A designated post requirement
for some types of offices was also adopted in the same
legislation.
97. During the spring of 1964, Donald L. Hollowell, a
black attorney in Atlanta, filed to run for a superior
court judgeship in the Septem ber 1964 prim ary in
Fulton County, a county in which superior court judges
could win the primary nomination by a plurality of the
vote. However, by virtue of the adoption of the majority
vote law in June of 1964, the primary was held using a
m ajority vote requirem ent instead of the previously
required plurality.
98. The Voting Rights Act of 1965 had the effect of
suspending the State of Georgia’s use of the literacy test
requirement for voter eligibility, although the provision
s till appeared in A rtic le II , §1, P a ra . I l l of the
Constitution of the State of Georgia of 1976. The provi
sion was removed from Georgia’s 1983 Constitution.
99. A ccording to a 1968 R ep o rt of th e U.S.
Commission on Civil Rights, Political Participation, pp.
12-13 (May 1968), there were 1,124,415 whites regis
tered in the State of Georgia as of December 1962, and
there were 167,663 non-whites registered to vote in the
Georgia at the time. As of August 31, 1967, there were
1,443,730 w hites reg istered to vote in Georgia and
123
332,496 non-whites registered to vote. An authentic
copy of the relevant pages of this publication is Exhibit
U.S. 54.
100. By statute enacted in 1966, the State of Georgia
restricted a person from assisting more than one illiter
ate voter during an election. This s ta tu te was held
unconstitutional in Morris v. Fort-son, 261 F. Supp. 538,
541 (N.D. Ga. 1966).
101. In 1968, Georgia passed a purge law pertaining to
municipalities maintaining their own registration sys
tems. The law required the municipal registrar at speci
fied times to remove from the voter list electors who (1)
had neither voted during the previous three years nor
requested continuation of registration, (2) had been con
victed of a crime constituting a disqualification from vot
ing, (3) had been adjudicated insane, or (4) had died. In
the same year, Georgia passed a law perta in ing to
municipalities maintaining their own registration sys
tems, which required the establishm ent of a central
office, and such other places as the municipality might
designate, a t which voter registration could be con
ducted.
102. Courts have found racial bloc voting in Georgia in
certain elections in Bleckley County, Hall v. Holder, 955
F.2d 1563 (11th Cir. 1992), cert, granted; Burke County,
Lodge v. Buxton, 639 F.2d 1358, 1378 (5th Cir. 1981);
C arro ll County, C arrollton B ranch o f N A A C P v.
Stallings, 829 F.2d 1547, 1559 (11th Cir. 1987); Colquitt
County, Cross v. Baxter, 604 F.2d 875, 880 n.8 (5th Cir.
1979); Dougherty County, Paige v. Gray, 437 F.Supp.
137,158 (M.D.Ga. 1977); Fulton County, Pitts v. Busbee,
395 F.Supp. 35, 40 (N.D.Ga. 1975); Putnam County,
Bailey v. Vining, 514 F.Supp. 452, 461 (M.D.Ga. 1981);
and Wilkes County, Wilkes County, Georgia v. United
124
States, 450 F.Supp. 1171,1174 (D.D.C. 1978). Fulton and
DeKalb Counties. Busbee v. Smith, 549 F. Supp. 494,499
(D.D.C. 1982) (three-judge court), a ffd , 459 U.S. 1166
(1983).
103. In each of the following counties and municipali
ties, located in whole or in part in the present Eleventh
Congressional District, voting rights litigation against
the jurisdiction resulted in changes in the challenged
electoral system(s) and/or judicial findings of racial bloc
voting: Baldwin County, Boddy v. Hall, Civil Action No.
82-406 (M.D.Ga. 5/17/84, 6/7/84) (Abrams Exhibits 2 and
3); C ity of M illedgeville, N A A C P v. C ity o f
M illedgeville, Civil Action No. 83-145-01 (M.D.Ga.
6/30/83) (Abrams Exhibit 4); Burke County, Lodge v.
Buxton, 639 F.2d 1358, 1378 (5th Cir. 1981) a f fd sub
nom. Rogers v. Lodge, 458 U.S. 613 (1982); Effingham
County, Local Organization fo r Voters E qua lity v.
Conaway, Civil Action No. CV484-39 (4/16/84) (Abrams
Exhibit 5); Butts County, Brown v. Bailey, Civil Action
No. 84-223 (M.D.Ga. 1984) (Abrams Exhibit 6); Greene
County, Bacon v. Higdon, Civil Action No. 85-40-ATH
(M.D.Ga. 1/10/86) (Abrams Exhibit 7); Henry County,
Head v. Henry County Board o f Commissioners, Civil
Action No. C-79-2063A (N.D.Ga. 7/17/80) (Abrams
E xh ib it 8); Je fferson County, Tom lin v. Jefferson
County Board o f C om m issioners, Civil Action No.
CV683-23 (S.D.Ga. 9/29/83) (Abrams Exhibit 9); Jenkins
County, Greene v. Bragg, Civil Action No. 691-078
(S.D.Ga. 10/4/93) (Abrams Exhibit 10); Putnam County,
B ailey v. V in ing , C ivil A ction No. 76-199-MAC
(M.D.Ga. 4/5/82, 4/22/82) (Abrams Exhibits 11 and 12);
Putnam County, Bailey v. Vining, 514 F.Supp. 452, 461
(M.D.Ga. 1981); Richmond County, Hamilton v. Board of
Commissioners o f R ichmond County, Georgia, Civil
Action No. CV178-226 (S.D.Ga. 1984) (Abrams Exhibit
125
13); City of Augusta, United States v. City Council o f
Augusta, Georgia, Civil Action No. CV187-004 (S.D.Ga.
7/22/88) (Abrams Exhibit); Screven County, Culver v.
K rulic, Civil Action No. 484-139 (S.D.Ga. 11/6/84)
(Abrams Exhibit 15); Screven County, United States v.
Screven County, Georgia, Civil Action No. CV692-154
(S.D.Ga. 12/16/92) (Abrams Exhibit 30); Twiggs County,
Bond v. White, 377 F.Supp. 514 (M.D.Ga. 1974); Wilkes
County, Wilkes County, Georgia v. United States, 450
F.Supp. 1171, 1174 (D.D.C. 1978); City of Waynesboro,
S u lliva n v. DeLoach, Civil Action No. CV176-238
(S.D.Ga. 9/22/77) (A bram s E x h ib it 16); C ity of
Warrenton, Warren County Branch o f the N AAC P v.
Haywood, Civil Action No. CV-187167 (S.D.Ga. 8/10/89)
(Abrams Exhibit 17).
SEGREGATED EDUCATIONAL INSTITU TIO NS
104. The 1945 Amendment to the Constitution of the
State of Georgia continued, inter alia, the then existing
requ irem en t of seg rega ted public schools and the
Georgia General Assembly enacted legislation requiring
segregated common schools and providing, among other
things, that “no teacher receiving or teaching white and
colored pupils in the same schools shall be allowed any
compensation out of the common school fund.” Both of
these provisions were subsequently abolished.
105. In 1945, the Georgia General Assembly passed a
proposed C onstitu tional Am endm ent perta in ing to
endowments to educational institutions open to the pub
lic. The proposed am endm ent s ta ted , among o ther
things, “that all endowments to institutions established
for white people, shall be limited to white people, and all
endowments to institutions established for colored peo
ple shall be limited to colored people.” These provisions
are no longer in force in Georgia.
126
106. The Suprem e C ourt in B row n v. B oard o f
Education, 347 U.S. 483 (1954), held that segregation of
white and Negro children in the public schools of a state
solely on the basis of their race, denies to Negro chil
dren the equal protection of the laws guaranteed by the
Fourteenth Amendment.
107. Follow ing the Suprem e C o u rt’s decision in
Brown v. Board o f Education, the Georgia General
Assembly in 1955 passed a resolution declaring that the
federal government had “greatly encroached, directly
and indirectly, upon the operation of State schools,” and
encouraged the United States Congress to convene a
constitutional convention for the purpose of amending
the United States Constitution to state that “nothing
shall, in any way, take away from, or in any manner
deprive, the right of the several States of the Union to
operate their school systems, and to have exclusive and
complete power to regulate and to determine the rules
and regulations for the conduct and operation of State
schools.” The United S tates Congress took no such
action.
108. Following the Supreme Court’s school desegrega
tion decisions in Brown v. Board o f Education, 347 U.S.
483 (1954) and Brown v. Board o f Education (No. II),
349 U.S. 294 (1955), the Georgia General Assembly in
1956 passed a resolution which expressed the view that
G eorgia had n e ith e r “su rren d ered to the G eneral
Government its right to maintain racially separate pub
lic schools and other facilities,” nor agreed to such a pro
hibition when ratifying the Fourteenth Amendment,
and that the United States Supreme Court’s decisions
were “null, void and of no force or effect.” No legislation
implementing the subject matter of said resolution was
passed.
127
109. In 1955, the Georgia General Assembly enacted
legislation which, inter alia, prohibited the use of state
or local funds for integrated public schools in the State
and made it a felony to violate such prohibitions. These
provisions are no longer in force in Georgia.
110. In 1955, the Georgia General Assembly passed
two resolutions calling upon the United States Congress
to convene a Constitutional Convention to amend the
Constitution: (1) to vest exclusive authority to regulate,
administer and operate public schools in the States; and
(2) to eliminate federal court jurisdiction in any case
drawing into question the administration of state public
school systems. The United States Congress took no
such action.
111. In 1956, the Georgia General Assembly adopted a
resolution urging the Congressional delegation from
Georgia to oppose HR 7535, a federal education bill
which called for the denial of federal funds to states that
continued to segregate th e ir public schools, for the
stated reasons that the bill would result in increased
taxation and that it constituted an unacceptable federal
encroachment on state sovereignty. The bill nonetheless
passed.
112. In 1956, the Georgia General Assembly passed a
bill enabling the Governor to close public schools under
certain prescribed circumstances, and, among other
things, to provide students in any such closed school dis
tricts with state and local public grants to be used for
educational purposes. Those provisions are no longer in
force in Georgia.
113. In 1956, the Georgia General Assembly provided
for the cut-off of further state funds upon admission of
blacks to a sta te university. Those provisions are no
longer in force in Georgia.
128
114. In 1957, the Georgia General Assembly adopted
legislation providing the Governor with the authority to
suspend the operation of the state compulsory school
attendance law when necessary because of any riot,
insurrection, public disorder, disturbance of the peace,
natural calamity or disaster, in order to protect persons
and property or to preserve the health and welfare of
the citizens, or to preserve the general welfare of the
state. These provisions are no longer in force in Georgia.
115. In 1957, the Georgia General Assembly passed a
resolution which expressed the view that the Congress
which had proposed the 14th and 15th Amendments was
improperly constituted, that the amendments were not
p rom ulgated in the m anner specified by the
Constitution, that ratification of those amendments had
been militarily coerced, and urged the United States
Congress to declare the F o u rteen th and F ifteen th
Amendments of the United States Constitution “null
and void and of no effect.” The United States Congress
took no such action.
116. In 1957, when the Judiciary Committee of the
United States House of Representatives was consider
ing civil r ig h ts leg isla tion , the G eorgia G eneral
Assembly passed a resolution requesting the United
States House of Representatives to permit “appropriate
public officials and c itizens to ap p ea r before the
[Judiciary] committee . . . ” In support of its request, the
General Assembly stated, among other things, that “the
[Judiciary] committee should not take speedy action on
legislation of this type, particularly in view of the fact
that there are serious legal and constitutional objections
thereto.”
117. In 1958, after President Eisenhower called out
m ilitary forces in L ittle Rock, A rkansas to enforce
129
school desegregation orders issued by the courts, the
Georgia General Assembly passed a resolution detailing
the view that the president had thereby exceeded his
powers, and that his actions were unlawful and uncon
stitutional for various reasons expressed in the resolu
tion. The resolution censured President Eisenhower for
those actions, and charged among other things that “the
president sacrificed the honesty and integrity of our
highest executive office on the alter of political expedi
ency to appease the NAACP and other radical, commu
nist sympathizing organizations.”
118. In 1959, the Georgia General Assembly passed a
resolution requesting that the Georgia Department of
Education prepare an outline of the grades and courses
taught in public schools in order “to offer a teaching
guide around which to build private schools in Georgia.”
The General Assembly declared that this effort was “for
the use of existing public and private schools, and for
the information of any and all other citizens of Georgia,”
and stated that “such information will be of the utmost
value in the assisting and solving of the most pressing
and grave problem ever to be presented to public educa
tion in this S ta te .. .”
119. In 1959, the Georgia General Assembly autho
rized the Governor to close any public school if the
Governor determ ined th a t continued operation was
likely to result in or cause violence or public disorder, or
that it was necessary to close the school in order to pre
serve order, peace and dignity of the state, and estab
lished criminal sanctions for failure to comply with such
an executive order. These provisions are no longer in
force in Georgia.
120. In 1959, the Georgia General Assembly autho
rized the Governor to close any branch or department
130
under the control of the Board of R egen ts of the
University System of Georgia if the Governor de te r
mined that continued operation was likely to result in or
cause violence or public disorder, or that it was neces
sary to close the school in order to preserve order, peace
and dignity of the state. These provisions are no longer
in force in Georgia.
121. The University of Georgia refused to admit black
students until that practice was declared unlawful and
enjoined by a federal court in 1961. Holmes v. Danner,
191 F. Supp. 394 (M.D.Ga. 1961); Ward v. Regents o f
University Svstem o f Georgia, 191 F. Supp. 491 (N.D.Ga.
1957).
122. The first black student is believed to have matric
ulated to the University of Georgia Law School in 1963;
he graduated in 1966.
123. From 1968 through 1973 school years, no first
year class of the University of Georgia law school had as
many as ten black students. Fourteen black students
were enrolled at the University of Georgia law school in
1973.
124. In 1959, the Georgia General Assembly passed a
resolution urging the United States Congress to call a
constitutional convention to amend the United States
Constitution to vest exclusive power over the operation
of public schools to the states and its subdivisions. The
United States Congress took no such action.
125. In 1960 the Georgia General Assembly enacted a
law which by its term s prohibited municipal corpora
tions in Georgia from levying taxes to support and main
tain integrated public schools. These provisions are no
longer in force in Georgia.
126. In 1961, the Georgia General Assembly enacted
131
legislation that established public grants to enable stu
dents to attend private schools of their choice in lieu of
public schools. One of the stated purposes of the legisla
tion was “to further secure the constitutional rights of
school children to attend private schools of their choice
in lieu of public schools.” These provisions are no longer
in force in Georgia.
127. In 1961, the Georgia General Assembly proposed
a s ta te co n stitu tio n a l am endm ent declaring th a t
“[flreedom from compulsory association at all levels of
public education shall be preserved inviolate,” and pro
viding for taxation to provide adequate education for
the citizens of Georgia.
128. In 1965, the Georgia General Assembly passed a
resolution requesting Congress to call a constitutional
convention to amend the United States Constitution,
proposing an amendment reading: “Among the rights
reserved to the states shall be the right to sole, and
exclusive jurisdiction of public school systems in the
separate states . . .” The United States Congress took
no such action.
129. In 1971, the House of the G eorgia G eneral
Assem bly passed a reso lu tion urg ing the Georgia
Congressional delegation to enact federal legislation
eliminating busing as a measure to achieve “an arbitrary
racial balance of pupil population,” for the stated reason
that busing is disruptive of the educational process, con
sumes a significant portion of the transported student’s
time which could be more productively used in pursuit
of educational and recreational activities, and diverts
school system assets from educational processes. The
United States Congress took no such action.
132
OTHER GEORGIA R A C IA L H ISTORY
130. In 1955, the Georgia General Assembly adopted a
resolution calling upon the Congress to convene a con
stitu tional convention to amend the U nited S ta tes
Constitution to permit members of the armed services
to serve in racially segregated units. No action was ever
taken by the Congress on the resolution.
131. In 1956, the Georgia General Assembly enacted a
law authorizing the General State Patrol and Georgia
Bureau of Investigation to enter in any county or munic
ipality upon the request of any citizen or official thereof,
to enforce Georgia’s segregation laws. These provisions
are no longer in force in Georgia.
132. In 1956, the General Assembly authorized state
and local governments to dispose of any property com
prising parks, swimming pools, or other property “dedi
cated to public use or recreation or park purposes,”
enacted a law requiring common carriers to provide sep
arate station rooms labelled “W hite W aiting Room,
Intrastate Passengers” and “Waiting Room, Interstate
Passengers and Colored In trasta te Passengers,” and
adopted an Act m aking it a m isdem eanor for an
intrastate traveler to occupy a waiting area not desig
nated for such traveler’s race. These provisions are no
longer in force in Georgia.
133. In 1956, the Georgia State Law Department pub
lished a “Compilation of Georgia Laws and Opinions of
the A ttorney General relating to Segretation of the
Races.” In his introduction to th is publication, the
Georgia A ttorney General stated , “A perusal of the
many segregation laws included herein, affecting as
they do almost every phase of human endeavor to which
the authority of government traditionally extends, con
vincingly demonstrates the predominant influence that
133
racial segregation has exerted in the sociological devel
opment of the two races in this State.” Georgia State
Law Departm ent, Compilation o f Georgia Laws and
O pinions o f the A tto rn e y G eneral re la tin g to
Secregation o f the Races, pp. 2, 3 (April 1956). An
authentic copy of this compilation of segregation laws is
Exhibit U.S. 55.
134. Prior to 1968, the State of Georgia maintained
racially segregated penal institutions. These segregated
penal institutions were declared unlawful in Wilson v.
Kelley, 294 F. Supp. 1005 (N.D. Ga. 1968), a ffd 393 U.S.
266 (1968)
GEORGIA SOCIOECONOMIC DEMOGRAPHICS
135. In Georgia, white households have a per capita
income of $15,832, black households have a per capita
income of $7,997, or 50.5 percen t of the per capita
income of w hites. 1990 Census, Table 17, Selected
Characteristics of Persons by Race and Hispanic Origin,
at p. 49. Abrams Exhibit 1.
136. In Georgia, white households have a mean income
of $41,043, black households have a mean income of
$23,752, or 57.9 percent of the mean income of whites.
1990 Census, Table 17, at p. 49. Abrams Exhibit 1.
137. In Georgia, of 1,757,768 white households, 90,861
(5.2%) have per capita income less than $5,000; of
573.290 black households, 94,391 (16.5%) have per capita
income less than $5,000. 1990 Census, Table 17, at p. 49.
Abrams Exhibit 1.
138. In Georgia, of 1,757,768 white households, 84,453
(4.8%) have per capita income of $100,000 or more; of
573.290 black households, 4,228 (0.7%) have per capita
income of $100,000 or more. 1990 Census, Table 17, at p.
49. Abrams Exhibit 1.
134
139. In Georgia, of 2,997,958 white persons 25 years
and over, 74.9 percent have high school diplomas or
higher education, and 21.8 percent have bachelor’s
degree or higher education; of 954,209 black persons 25
years and over, 58.6 percent have high school diplomas
or higher education and 11.0 have bachelor’s or higher
education. 1990 Census, Table 17, a t p. 49. Abrams
Exhibit 1.
140. In Georgia, of 2,479,480 white persons 16 years
and over in the labor force, 98,944 (4.0%) are unem
ployed; of 802,174 black persons 16 years and over in the
labor force, 85,049 (10.6%) a re unem ployed. 1990
Census, Table 17, at p. 49. Abrams Exhibit 1.
141. In Georgia, of 1,279,452 white families, 79,248
(6.2%) have income below poverty level; of 418,583 black
families, 114,519 (27.4%) have income below poverty
level. 1990 Census, Table 17, at p. 50. Abrams Exhibit 1.
142. In Georgia, of 4,496,697 white persons, 394,561
(8.8%) have income below poverty level; of 1,678,436
black persons, 507,992 (30.3%) have income below
poverty level 1990 Census, Table 17, at p. 50. Abrams
Exhibit 1.
143. In Georgia, of 313,042 white children under age
five, 34,251 (10.9%) have income below poverty level; of
162,216 black children under age five, 70,953 (43.7%)
have income below poverty level. 1990 Census, Table 17,
at p. 50. Abrams Exhibit 1.
144. In Georgia, of 183,179 white persons 75 and over,
38,947 (21.3%) have income below poverty level; of
51,383 black persons 75 and over, 23,605 (45.9%) have
income below poverty level. 1990 Census, Table 17, at p.
50. Abrams Exhibit 1.
145. In Georgia, of 1,279,452 white families, 139,156
135
(10.9%) are female-headed households with no husband
present; of 418,583 black families, 180,908 (43.2%) are
female-headed households with no husband present.
1990 Census, Table 17, at p. 50. Abrams Exhibit 1.
1970s RED ISTRICTING
156. On November 4, 1971, Georgia submitted a con
gressional redistricting plan to the Attorney General for
review pursuant to Section 5 of the Voting Rights Act.
The plan had been enacted into law in October 1971. See
Exhibit U.S. 1.
157. According to the state’s November 4, 1971 sub
mission of this redistricting plan was to comply with the
one person, one vote principle. “The act was adopted in
order to ensure that the apportionment of the Georgia
Congressional Districts corresponds to the one man-one
vote standard enunciated by the United States Supreme
Court.” 1971 submission, at 1.
158. The 1971 redistricting was required by an unre
ported three-judge court opinion and order in Cahoon v.
Maddox, C.A. No. 13011 (N.D. Ga., January 12,1970).
159. Changes in the distribution of the state’s popula
tion between 1960 and 1970 made it necessary to change
the boundaries between the various congressional dis
tricts. According to the state’s November 4, 1971, sub
mission, “The difference between the old and new law is
tha t the new apportionment act establishes districts
which are v e ry n early equal in popu la tion .” 1971
Submission, at 1.
160. According to data compiled by the U.S. Bureau of
the Census, the nonwhite population of Fulton County,
Georgia, increased from 34.8 percent of the county’s
total population in 1960 to 39.3 percent in 1970.
136
161. According to data compiled by the U.S. Bureau of
the Census, the nonwhite population of the city of
Atlanta (located primarily in Fulton County but with a
small portion in DeKalb County) increased from 38.3
percent in 1960 to 51.6 percent in 1970.
162. On the basis of the one-person, one-vote stand
ard, the ideal congressional d istrict population after
1970 was 458,958. T hat p a rt of the city of A tlan ta
located in Fulton County had a population of 450,286 in
1970, only 8,672 persons less than the size of an ideal
congressional district.
163. In its November 4, 1971, submission the State of
Georgia stated that it used racially neutral redistricting
principles in creating its congressional redistricting
plan. Attorney General Arthur K. Bolton stated in the
submission that racial composition data “were not used
by the General Assembly to calculate the new districts.”
The state did not include such data with its Section 5
submission, noting that “population information by race
may be obtained from the Bureau of the Census.” 1971
Submission, pp. 2, 5.
164. The state’s 1971 submission stated: “The General
Assembly contracted with the computer center at the
University of Georgia to assist in establishing equal dis
tricts. The contract specifically prohibited the Center
from programming racial data.” 1971 Submission, at 3.
165. The sta te’s 1971 submission included, as docu
mentation of the contract referenced in the preceding
paragraph, a new spaper article about the contract:
Prentice Palmer, “COMPUTER IN REMAP TO SKIP
VOTE RECORD: RACE MAKEUP ALSO RULED
OUT AS HOUSE HIRES ATHENS TEAM,” Atlanta
Journal, April 8, 1971, at 1A, 6A. An authentic copy of
the newspaper article referenced by the state is Exhibit
U.S. 2.
137
166. According to the state’s 1971 submission, one rea
son why the state prohibited the computer center from
entering racial data was that “it was believed that the
use of such data might itself be constitutionally imper
missible.” 1971 Submission, at 3.
167. According to the sta te’s 1971 submission, “the
basic plan was to maintain the integrity of the counties,
where possible. W here it was necessary to divide a
county, the breakdown was according to census county
division, census tract, enumerated district, or in a few
instances, census blocks.” 1971 Submission, p. 3.
168. The existing congressional redistricting plan, in
effect since 1964, split Fulton County betw een the
fourth and fifth congressional districts by assigning a
strip along the eastern border of Fulton County to the
fourth district. An authentic copy of the map of this
plan, as included in the state’s November 4, 1971 sub
mission is Exhibit U.S. 3.
169. The congressional redistricting plan adopted by
the state in 1971, restored most of the aforesaid portion
of eastern Fulton County to the Fifth Congressional
D istrict and assigned in portions of southern Fulton
County to the F ourth and the Sixth Congressional
Districts. An authentic copy of the map of this plan, as
included in the state’s November 4, 1971 submission is
Exhibit U.S. 4.
170. The 1971 congressional redistricting plan split
F u lton C ounty in to the F o u rth , F if th and S ix th
Districts. One county was split between the Seventh,
and Ninth Congressional Districts.
171. The congressional redistricting plan adopted by
the state in October 1971 assigned portions of south
western A tlanta to the Sixth Congressional District.
138
See Exhibit U.S. 5.
172. According to the aforesaid comm ent of the
Fulton County Democratic party, both A tlanta Vice-
Mayor Maynard Jackson, who is black, and Andrew
Young, a black candidate for the fifth district seat in
1970 against incumbent Republican Fletcher Thompson,
resided in the area of southwestern Atlanta moved from
the Fifth to the Sixth congressional district in the plan
adopted by the state in 1971. Exhibit U.S. 5.
173. In those portions of the city of A tlanta moved
from the Fifth to the Sixth Congressional District, 8,907
(52.6 percent) of the registered voters were black. In
those areas of Fulton County outside the city limits of
A tlan ta th a t rem ained in th e F ifth C ongressional
District, 56,588 (98.9 percent) of the registered voters
were white. Exhibit U.S. 5.
174. According to E xhibit U.S. 5, Fulton County
Democratic Party incumbent Congressman Fletcher
Thompson and state Senator Frank Coggins, one of the
co-authors of the congressional red is tr ic tin g plan
adopted by the state in October 1971, resided in that
area of Fulton County south of Atlanta that remained in
the Fifth Congressional District.
175. Senator Leroy Johnson, the first black person
elected to the Georgia General Assembly in the twenti
eth century, developed a partial congressional redistrict
ing plan in which the Fifth Congressional District would
be composed primarily of the city of Atlanta, with some
predominantly white areas of northern Fulton County
shifted into the fourth district. The vote in the Senate
on the Johnson plan on September 30, 1971, was nine in
favor, with 43 opposed. Exhibit U.S. 6.
176. Representative Benjamin D. Brown introduced a
139
partial congressional redistricting plan in the Georgia
House on October 4, 1971. Using 1970 census data, the
nonwhite population of the Fifth District was 49.2 per
cent in the Brown plan. The Brown plan placed predom
inantly white northern portions of Fulton County in the
Fourth Congressional District. Exhibit U.S. 6.
177. The congressional redistricting plans submitted
by Senator Johnson and Representative Brown, as ref
erenced in the preceding paragraphs, were defeated by
substantial margins. However, all 13 black representa
tives in the House voted for the Brown plan, and both
black senators voted for the Johnson plan. 1971 Senate
Journal, at 118; 1971 House Journal, at 3611-3612.
178. In the House Com mittee on Legislative and
Congressional Reapportionment, Representative Grace
Hamilton, the only black member of the committee,
introduced a congressional redistricting plan, which was
not approved by the committee. Exhibit U.S. 6.
179. On February 11, 1972, the Attorney General of
the United States interposed a timely objection to the
congressional redistricting plan submitted on November
4, 1971, on the grounds that the state had not met its
burden of proof regarding the boundaries between the
Fifth and Sixth Congressional Districts. “With respect
to those [districts], after careful review of all the infor
mation available to us we have been unable to conclude,
as we must under the Voting Rights Act, that these new
boundaries will not have a discriminatory racial effect
on vo ting by m inim izing or d ilu ting black voting
strength in the Atlanta area.” Exhibit U.S. 7.
180. In 1982, a federal court concluded that the con
gressional redistricting plan adopted by the state of
Georgia in 1971 was drawn to fragment black voting
strength in the Atlanta area and to place the residences
140
of black persons clearly recognized as potential candi
dates for election to Congress from the Fifth District
ju s t across the bo rder in the S ixth C ongressional
District. Busbee v. Smith, 549 F. Supp. 494, 500 (D.D.C.
1982).
181. The United States Attorney General precleared a
revised congressional redistricting plan for the State of
Georgia on April 11,1972. Exhibit U.S. 8.
241. In 1972, Andrew Young became the first black
person since the Reconstruction period to be elected to
the United States Congress from the State of Georgia.
From 1972 until 1992, no more than one black person at
a time was a member of Georgia’s congressional delega
tion.
141
I. DECLARATIONS OF SELWYN CARTER
[Abrams’ Exh. 35]
IN THE UNITED STATES DISTRICT COURT
SOUTHERN DISTRICT OF GEORGIA
AUGUSTA DIVISION
Civil Action No.
CV 194-008
David a J ohnson, et al . Plaintiffs,
v.
Zell Mil l e r , et al., Defendants,
and
Lucious Abrams, J r ., et al ,
Defendant-Intervenors,
U nited States of America ,
Defendant-Intervenor.
D ECLARATION OF SELW YN CARTER
Pursuant to 28 U.S.C. §1746,1 declare that:
1. My name is Selwyn Carter. I am the Director of the
Voting R ights P rogram of the Southern Regional
Council, Atlanta, Georgia. The Council is a non-profit
human rights organization.
2. I am a g raduate of City College of New York,
where I majored in economics. I have completed the
course work for a masters degree in political manage
ment at the Graduate School of Political Management at
George Washington University.
142
3. The Voting R ights P rog ram of th e S ou thern
Regional Council provides various services to commu
nity groups and sub-units of government, primarily in
the South. These services include preparing proposed
redistrieting plans for local and state legislative bodies
and for the Congress. I have prepared numerous re
districting plans using redistricting computer software
and have given declarations similar to this in other court
cases.
4. During July, 1995 I was contacted by Laughlin
McDonald, one of the attorneys for the Abrams inter
veners. He asked me to prepare one or more congres
sional redistricting plans for the state of Georgia. He
advised me that the Supreme Court invalidated the
existing Eleventh District on June 29, 1995 in Miller v.
Johnson, and that the Court concluded that race had
been the predominant factor in the 1992 redistricting
process and that the state had subordinated all its tradi
tional redistricting principles to race.
5. In preparing my plans, Mr. McDonald instructed me
to rely to the greatest extent possible upon the state’s
traditional non-racial redistricting principles, which I
attem pted to do. In this connection, he provided me
with a copy of the redistricting guidelines adopted by
the state during the 1991-1992 redistricting process. A
copy of these guidelines is attached to this declaration,
6. Mr. McDonald also instructed me that while race
could be considered as one of many factors in redistrict
ing, race could not be the predominant factor in the con
struction of any redistricting plans nor could the state’s
traditional redistricting principles be subordinated to
race.
7. In preparing the redistricting plans my staff and I
applied the following criteria:
143
a. Complying with one person-one vote as nearly as is
practicable.
b. Using only single-member districts.
c. Insuring that districts are composed of contiguous
geography.
d. Maintaining the integrity of political subdivisions
as much as possible.
e. Maintaining the cores of existing districts.
f. Protecting and avoiding contests between incum
bents.
g. Using local voting precinct boundary lines as the
basic district building blocks, or where that is not practi
cable census block group geography.
8. Mr. McDonald further advised me that the plans he
wanted me to draw were intended to be remedial and to
be proposed remedies for the violation found by the
Supreme Court, which he said involved the construction
of the Eleventh District. He asked me to focus on the
Eleventh District and not to change other districts any
more than was necessary in redrawing the Eleventh and
in complying with the sta te’s traditional redistricting
principles.
9. Mr. McDonald also advised me that the plans he
wanted me to draw might be filed with the court as pro
posed court ordered rem edies, and th a t as a conse
quence I should try to keep the population deviations as
low as possible.
10. Applying the instructions and guidelines described
above, my staff and I prepared three redistricting plans.
The f ir s t is d a ted A ugust 8, 1995 and is m arked
SRC/ACLU Plan A. A copy of Plan A is attached. This
plan was actually introduced before the house and sen-
144
ate reapportionment committees on behalf of the black
legislative caucus but was not adopted. Under Plan A,
the Eleventh District consists of the following whole
counties: B urke, Jefferson , W ashington, Hancock,
T aliaferro , G reene, Putnam , Jasp e r, N ew ton, and
Rockdale. Portions of Richmond, Warren, DeKalb and
Fulton Counties are also included in the E leven th
District. Under the pre-existing 1992 plan invalidated
by the Supreme Court nine counties were split. The
total deviation among districts under Plan A is 0.21%.
11. In constructing Plan A, I avoided using so-called
land b ridges and did not include any p o rtions of
Savannah. I also a ttem pted to make the E leven th
District as regular as possible while at the same time
applying the guidelines described above. Race was not
the predominant factor in the construction of Plan A,
nor were the state’s traditional redistricting principles
subordinated to race. That is apparent from: (1) the reg
ular shape and compactness of the Eleventh District; (2)
the number of whole counties in the district; (3) the fact
that the black voting age population in the district was
reduced from 60.4% to 53.05%; (4) the fact that the devi
ation is only 0.29%; (5) the fact that the district follows
the core of the old Tenth District under the 1982 plan;
and, (6) insofar as I am aware no two incumbents are
placed in the same district.
12. My staff and I drew a second redistricting plan
dated August 9,1995 which is marked SRC/ACLU Plan
B. A copy of Plan B is attached. The Eleventh District
under Plan B has the same configuration as under Plan
A. The difference between the two plans is that fewer
counties are split in Plan B. I drew Plan B simply to
show that a plan could be drawn, in conformity with the
s ta te ’s traditional redistricting principles, that mini
mized the number of split counties. However, because
145
fewer counties are split under Plan B it has a deviation
of 0.51%, which is higher than the deviation in Plan A.
13. My staff and I drew a th ird redistricting plan
which is marked Plan C, a copy of which is attached.
Plan C was generated using the state’s, rather than the
Southern Regional Council’s, computer. The significant
differences in Plan C and the other two plans are that no
portions of Richmond County are included in the
E leventh D istrict, and only two counties are split,
Fulton and DeKalb. Overall, only seven counties are
split in Plan C, Muscogee, Fulton, DeKalb, Clayton,
Cobb, Gwinnett, and Bibb, all of which are in heavily
populated urban areas of the state. However, because
few counties were split, the total deviation among dis
tricts is 0.99%, the greatest deviation among the three
plans.
I declare under penalty of perjury that the above and
foregoing is true and correct. Executed a t A tlanta,
Georgia this 31st day of August, 1995.
Selwyn Carter
8/31/95
146
IN THE UNITED STATES DISTRICT COURT
SOUTHERN DISTRICT OF GEORGIA
AUGUSTA DIVISION
Civil Action No.
CV 194-008
David a J ohnson, et a l . Plaintiffs,
v.
Zell Mil l e r , et al ., Defendants,
and
Lucious Abrams, J r ., et al ,
Defendant-Intervenors,
U nited States of A merica
Defendant-Intervenor.
SUPPLEM ENTAL DECLARATION OF
SELW YN CARTER
Pursuant to 28 U.S.C. §1746,1 declare that:
1. My name is Selwyn Carter. The purpose of this sup
plemental declaration is to correct an error in my prior
declaration.
2. The data for Plan A, which was dated August 8,
1995, included in my prior declaration was for a revision
of Plan A and was mistakenly attached to the declara
tion. The correct demographic data and map for Plan A
is attached and is dated July 31, 1995. The differences
between the July 31, 1995 plan and the August 8, 1995
147
revision are minor, the principle one being that the July
31, 1995 plan has a to ta l deviation of .29% and the
August 8, 1995 plan has a total deviation of .45%. The
configuration of the Eleventh District is the same under
both plans.
I declare under penalty of perjury that the above and
foregoing is true and correct. Executed at A tlanta,
Georgia this day of October, 1995.
Selwin Carter
10/10/95
148
IN THE UNITED STATES DISTRICT COURT
SOUTHERN DISTRICT OF GEORGIA
AUGUSTA DIVISION
Civil Action No.
CV 194-008
Davida J ohnson, et al . Plaintiffs,
v.
Zell Mil l e r , et al ., Defendants,
and
Lucious Abrams, Jr., e t al ,
Defendant-Intervenors,
U nited States of A m erica ,
Defendant Intervenor.
SECOND SUPPLEM ENTAL D ECLARATION OF
SELW YN CARTER
Pursuant to 28 U.S.C. §1746,1 declare that:
1. My name is Selwyn Carter.
2. I was asked by Laughlin McDonald, one of the
attorneys for the Abrams intervenors, to prepare a con
gressional red istricting plan tha t complied with the
Order of the Court dated October 17, 1995, i.e., that
“makes the least changes, in terms of line drawing, in
Georgia’s present congressional plan but at the same
time brings the Eleventh District into compliance with
the United States Constitution.” I was also asked by Mr.
149
McDonald to draw a plan applying the same standards
set out above that assumed that the Second District is
now unconstitutional.
3. In addition, Mr. McDonald requested me to draw a
plan that complied with the subsequent Order of the
Court dated October 20, 1995, i.e., that was “based on
the first plan that Georgia submitted to the Department
of Justice for preclearance.”
4. Pursuant to these instruction a plan was prepared
under my supervision, a copy of which is attached and
designated ACLU1A. This plan assum es th a t the
Second District is also unconstitutional.
5. In drawing ACLU1A, I a ttem pted to make the
least changes possible in the existing plan but at the
same time address the problem areas identified in the
opinions of this Court and the Supreme Court.
6. The N in th D is tr ic t is to ta lly unchanged in
ACLU1A from the existing plan.
7. There are only minimal changes in the Fourth,
Fifth, Sixth, Seventh, and Tenth Districts compared to
the existing plan. In general the changes th a t were
made were necessary to comply with one person, one
vote, and because of changes made in other districts,
primarily in the Second and Eleventh Districts.
8. More specifically, in the Tenth District the changes
were limited to precinct lines in Richmond and Gwinnett
Counties. The changes in Richmond County were made
to respond to the finding of the Court that the configura
tion of the existing plan was designed only to include or
link black neighborhoods in the E leventh D istrict.
Under ACLU1A, the Eleventh District is not designed
simply to link black neighborhoods, but includes a sub
stantial white population. In addition, the split was
150
made to comply w ith the w ishes of th e Richmond
County Commission that the county be split to maxi
mize the representation of the county in Congress. Race
was therefore not the predominant reason for splitting
Richmond County in ACLU1A, nor is the split simply
along racial lines. The E leven th D is tric t includes
115,759 persons from Richmond County of whom 50.19%
are black and are of voting age.
9. The changes in the Tenth District in Gwinnett were
made to comply with the one person, one vote standard,
to compensate for the previously described changes in
the Tenth in Richmond County, and to maintain a mini
mum of county splits.
10. In the Sixth District, precinct changes were lim
ited to Cobb and Gwinnett Counties. The changes in the
Sixth in Gwinnett were the result of the changes in
Gwinnett in the Tenth described above. The changes in
Cobb County were limited to two precincts and were
caused by the changes in Gwinnett and because of the
need to comply with one person, one vote.
11. In the Seventh District, the changes were the
result of the two precinct changes in Cobb County in the
Sixth described above. Secondly, M eriwether County
was included in the Seventh because the Courts criti
cized its inclusion into the Second. Further, its inclusion
into the Seventh was to comply with one person, one
vote.
11. The changes to the Eighth District resulted from
bringing the Second and the Eleventh D istricts into
compliance with the Courts’ standards as described
above. The removal of M eriwether County from the
Second District and population changes resulting from
the elimination of splits in the counties of Lowndes,
C olquitt, D ougherty, Lee, C risp, Dooly, H ouston,
151
Bleckley, Twiggs and Crawford necessitated changes to
the Eight and Third Districts to bring them into compli
ance with one person, one vote. While there are many
ways to insure compliance with one person, one vote,
these particular changes were made to achieve mini
mum disruption to the existing plan.
12. The changes to the F irst Congressional District
resulted from the removal of the Chatham/Effingham
area from the Eleventh District.
13. No counties were split in the F irst or the Eight
Districts.
14. While ACLU1A is based specifically on the exist
ing plan rather than the first plan enacted by the state
and submitted to the Department of Justice, ACLU1A
fairly incorporates significant features of the first plan.
The overriding concern in the construction of ACLU1A
was to correct the defects identified by the Courts in the
existing plan.
I declare under penalty of perjury that the above and
foregoing is true and correct. Executed at A tlanta,
Georgia this day of October, 1995.
Selwyn Carter
October 25,1995
152
J. TESTIMONY OF LINDA MEGGERS
[Excerpts from Hearing Testimony, 10/30/95-10/31/95:
Tr. p. 20, line 1-p. 21, line 14; p. 26, lines 2-20; p. 31, line
17-p. 32, line 8]
MR. PARKS: Your Honors, we are going to offer Ms.
Meggers as a political analyst and expert in redistrict
ing. As to her qualifications I would point out that both
this court and the Supreme Court have recognized her
as a most knowledgable person in this area.
I intend to ask her both factual questions and ques
tions that will require her to give her opinion. I would
like to have her qualified as an expert.
JUDGE BOWEN: Considering the findings of this
court and the Supreme Court it would be difficult to
deny Ms. Meggers the right to state her opinions. And
I’m confident, Ms. Meggers, if you are ask[ed] to state
any opinion you do not feel comfortable with you will let
is know.
You may elicit her opinion from time to time as is nec
essary within the sphere of her expertise.
MR. PARKS: Thank you, your Honor.
Q. Ms. Meggers, were you involved in the drawing of
the Second Congressional District?
A. Yes.
Q. W hen, in re la tionsh ip to the draw ing of the
Eleventh Congressional District, was the idea or deci
sion made to try to draw a majority black district south
w est Georgia[;] did th a t come before or a fte r [the]
Eleventh District?
A. It came after.
153
Q. In what context?
A. I think at the outset of 1991 there was a purpose
ful effort on behalf of the General Assembly to draw a
second minority majority district. The creation of a the
Second as a m inority m ajo rity d is tr ic t is c learly
reflected in the second plan produced by the General
Assembly.
To answ er your question creation of the Second
Congressional District as a minority majority district
came later.
Q. Was the portion of Bibb County that was placed in
the Eleventh District primarily, in the first plan, the
predominantly black population concentrations of Macon
and Bibb County?
A. Yes, they were the major black precincts.
Q. The Eleventh District ending here is what we are
seeing her[e] the result of what is known as the Macon
Savannah trade?
A. Yes, on the third try at drawing the Congressional
plan Bibb County was rem oved from the E leventh
District and that portion of the black population previ
ously in the Eleventh was attached to Second District.
Q. For what reason?
A. That was the one way in which we could achieve a
m ajo rity black VAP in th e Second C ongressional
District.
Q. So it was racially motivated?
A. Yes.
jjs He Hi Hi ^
154
Q. F ir s t tak in g M uscogee C ounty w here has
Muscogee County been located, traditionally?
A. In the Third Congressional District.
Q. Why was that?
A. The Third Congressional District, up until 1991,
was what we had dubbed, basically, the military district.
It included Houston with Warner Robins and Muscogee
with Fort Benning.
Q. Where is Fort Benning located?
A. Fort Benning starts, on this one here, it s ta rts
here and takes in a big part of Chattahoochee County. It
comes down like this and the bulk of Chattahoochee
County is below it.
Q. So it tak e s in th e n o rth e rn portion [of]
Chattahoochee County and the southern portion of
Muscogee?
A. Yes.
155
K. TESTIMONY OF SANFORD BISHOP
[Excerpt from Hearing Testimony, 10/30/95-10/31/95 Tr.
p. 114, line 17-p. 116, line 7]
During the reapportionm ent process whenever the
question of splitting counties or cities was raised people
in the those areas would be on opposite sides. Some will
say we don’t want our county or city split. So I felt as
we went into the special session it would be a good idea
to get an indication from the people in those cities or
counties as to how they felt about it. In talking with
members of city councils and mayors and the adminis
trators for housing authorities and school board mem
bers from across the Second District many of them said
we like things just the way they are. The only informa
tion I could give them was to write your legislator and
let him know how you feel because that’s going to be an
issue. If you are comfortable with this the way it is let
someone know. The m ayors and councils of Macon,
Columbus, Albany and Valdosta have passed resolutions
saying they were comfortable with the split. In light of
the redistricting process that was pending they did not
want making their counties whole to be a consideration.
They like having two members of Congress to have
their needs addressed. They felt in the long run it was
serving their best interests contrary to what they ini
tially felt.
Therefore, they were comfortable in passing a resolu
tion in Macon, Columbus and Albany and those resolu
tions were passed unanimously. In Valdosta it was by a
majority. It was a feeling, even in the business areas, in
Columbus, for example, where there had been a great
deal of consternation about splitting Muscogee County.
A fte r th e com bined sew er overflow s funds w ere
obtained as a result of the freshmen Congressmen work-
156
ing together to get $20 million for the project th a t
everybody th o u g h t was going to be im possible.
Working together one Democrat and one Republican
were able to get that money. When the Republicans on
the Senate side were going to take that money out the
Republican member from the Third district was able to
keep that money and we were able to secure that pro
ject for Columbus.
157
L. ARGUMENT. LAUGHLIN McDONALD
[Excerpts from Hearing Testimony, 10/30/95-10/31/95 Tr.
p. 277, lines 10-22; p. 279, line 15 to p. 280, line 7; p. 282,
lines 2-12; p. 283, lines 11-15; p. 283, line 21-p. 284,
line 10]
Now, the State argues that it was compelled by the
Department of Justice to draw a plan containing two
and then th ree m inority m ajority d istric ts, but the
record does not reflect that. Prior to any involvement by
the Department of Justice. Prior to the submission of
any plan by [sic] to the Attorney General the leadership
made the decision, based on the fact that the State was
gaining a congressional district, that they would draw a
second majority black district and that's what they did
in the first plan they drew draw [sic]. So it was the
State’s announced policy of having two majority black
districts.
* * *
I want to talk for a moment about the retrogression
standard and what serves as a basis for determining
that. I submit that the 1982 plan cannot serve as the
bench mark for retrogression for two reason[s]. F irst of
all that plan is and was an unconstitutional plan. It was
in violation of the one person one vote. The Second [sic]
major problem is that it contains a different number of
seats from any existing plan. It was a ten seat plan and
we are now dealing with an eleven seat plant;] it is like
comparing apples and oranges. Even if we assume,
despite the fact that it is unconstitutional, despite the
fact that it contains a different number of seats[,] that
[the] 1982 plan should be the bench mark[,] I think any
plan that contains less than ^wo minority majority dis
tricts is retrogressive for the simple reason that one out
of 11 is retrogressive when compared to one out of ten.
^ Sfc
158
The Court said, in an opinion that has been criticized
by some, that there are no diminimous [sic] deviations in
a congressional redistricting plan and invalidated the
New Jersey plan because the court said you could have
drawn one w ith a .4 percent deviation. There is no
diminimous [sic] deviation and a jurisdiction is obligated
to make a good faith effort to comply exactly with popu
lation equality concepts. Of course, if a legislature has to
do that I think it is even more incumbent on [a] District
Court to do that when drawing a plan.
Now, in our plan our deviation is .93. So I did want to
alert the Court that in our view that is something that
the Court needs to be concerned about, but it may be
removed by Carsher [sic] as long as you have an articu
lable basis for it.
In our plan we tried to split as few as possible because
we wanted to respond to the criticism, but it is clear
that you could split some counties here and cure those
deviations. It might well be that principle of maintaining
the integrity of those counties articulated, expressed
and justified in the record would excuse the deviations
of that .9 percent. I suppose we are suggesting that it
would because we are presenting this plan as one we
think would be acceptable. I think some of ones that go
up to 1.9 nine that’s just inviting a problem, it seems to
me, for the court. Aside from the retrogression plan I
think that’s a serious legal issue if the Court were to
draw a plan that contains less than two minority dis
tricts.
159
M. TESTIMONY OF SELWYN CARTER
[Excerpt from Hearing Testimony, 10/30/95-10/31/95 Tr.
p. 294, line 17-p. 295, line 3; p. 295, line 14-p. 296, line 12;
p. 297, line 12-p. 309, line 17; p. 310, line 14-p. 313, line 3;
p. 313, line 19-p. 323, line 21; p. 324-lines 1-25; p. 325, line
9-p. 326, line 15; p. 330, lines 4-12; p. 330, line 19-p. 331,
line 6; p. 332, lines 6-14; p. 333, lines 3-6; p. 338, lines 10-
20; p. 339, line 3-p. 340, line 19.]
Q. Would you outline your duties with the Southern
Regional Council?
A. Certainly. F irs t of all the council is the oldest
inter-racial organization in the southeast. We view our
selves as promoting democratic practices in govern
ment. My work is to oversee the council^ work in the
area of promoting fairness in redistricting across the
south. We assist sta te legislatures, attorney generals
and community leaders in drafting districting plans and
in assessing plans drafted by others for fairness and
compliance with the Voting Rights Act.
* * * *
MR. McDONALD: Your honor, we would like to have
Mr. Carter qualified as an expert in the field of demogra
phy and redistricting.
JUDGE BOWEN: You may elicit his opinions.
Q. Were you asked to draw or produce any congres
sional districting plans by the Abrams Intervenors?
A. Yes, I was.
Q. What did you do?
A. I developed a number of plans some of which were
p resen ted to the L eg isla tu re or ra th e r before the
160
Reapportionm ent Committee of the Legislature. A
number of them which were presented with affidavits
here to this Court.
My basic goal in preparing those plans was to try to
show it is possible to develop a constitutional redistrict
ing plan which is fair to both black and whites voters in
the State of Georgia and comply with this Court’s ruling
in Miller v Johnson. To show that it is possible to draw a
plan in which A frican A m erican v o te rs com prise
approximately 50 percent of the voting age population
of a district and at the same time show that race was not
a factor.
MR. McDONALD: Then, your Honor, we have a map
which I think it would make sense to introduce as a sep
arate exhibit which is the ACLU1A Plan.
JUDGE BOWEN: Do you want to call that 36?
MR. McDONALD: Yes, sir.
JUDGE BOWEN: All right. 36 is admitted. (Abrams
Intervenors Exhibit 36 was received in evidence.)
Q. Mr. Carter, I principally want to ask you about
what has been marked Abrams Intervenors Exhibit 36.
Did you either construct or supervise the drawing of
this plan ACLU1A?
A. The first three plans you mentioned I did draw my
myself. This particular plan was drawn under my super
vision.
Q. Was it drawn on the state computer?
A. That’s correct.
Q. What did you do in generating this plan?
161
A. The overriding methodology in developing this
plan was to correct the constitutional defects in the
Eleventh Congressional District and assume that the
Second Congressional District was unconstitutional and
correct to [sic] assumed defect in that district and to
maintain the remaining districts with as little change as
possible.
Q. W hat changes did you make to the E leventh
Congressional District compared to the existing plan?
The existing plan is below the plan that you prepared,
can you see that?
A. Yes, I can.
T here w ere a num ber of changes. The principal
changes were the extensions of the d istric t through
Effingham County and Chatham county were removed
because they were identified by th is Court and the
Supreme Court at [sic] being clear examples of racial
gerrymandering of the district.
In addition the way in which this district goes into
Richmond County through what’s been called the back
door was corrected. That was corrected by bringing in a
considerable number of white voters previously in the
Tenth District and transferring them to the Eleventh
District.
Q. Richmond County is split in your plan?
A. Yes, sir, it is split.
Q. Is that split made along racial lines?
A. No, it’s not. In fact, Richmond County contains
115,000 voters in the Eleventh district^] Approximately
50 percent African America[n]s of voting age and 47
percent white voters. I t is difficult to give the exact
number for the black and white voters because on the
162
plans that are produced on the system at the Southern
Regional Council we publish the demographics showing
the white population and the white voting age popula
tion and the black population and the black voting age
population and the system used by the State of Georgia
doesn’t.
Q. Why was this plan drawn on the State’s computer
as opposed to the Southern Regional Council’s com
puter?
A. The State’s computer contained the most current
precinct boundaries and we wanted to draw the plan
using the most up to date precinct boundaries so as not
to split precincts.
Q. What other changes did you make compared to the
existing plan?
A. In addition to the ones in Effinghamf,] Chatham
and Richmond there are a couple of areas where rural
counties had been split in the Eleventh District, the
Counties of Baldwin, Twiggs and Wilkes County. The
splits in those rural counties were eliminated in this par
ticular plan and they were primarily eliminated because
our goal was to essentially restore the integrity of those
rural political subdivisions and make the integrity of
those subdivisions predominant. Clearly here in [sic] an
attem pt was made to draw a plan in which race was not
the predominant factor. If race was predominant we
would have drawn this plan at a lower level of geogra
phy.
Additionally the other principal change has to do with
the border between the Eleventh and Fifth Districts.
C urren tly the F ifth D istric t ex tends from Fulton
County across the border into DeKalb County and into
what is known as east Atlanta. In ACLU1A the division
163
between the Fifth District and the Eleventh District is
the county line.
Q. How many whole counties are there in 1A?
A. Do you want me to get up there and count them?
Q. Are there 14?
A. There are probably 14 or 15.
C H IE F JU D G E E D E N F IE L D : W hat was your
question?
MR. McDONALD: I asked him how m any whole
counties w ere in th is version of th e E lev en th
Congressional District?
A. There are 14 counties in the d istric t th a t are
intact.
Q. Why did you include 14 whole counties in the
Eleventh?
A. As I previously mentioned one of the overriding
goals was to try to show that it is possible to draw dis
tricts in the State of Georgia for members of Congress
that did not use race as a predominant feature and on the
other hand were fair to black as well as white voters.
During the special session there was much fuss made
about m aintaining the in teg rity of county boundary
lines, particularly in the rural county areas and it was an
{sic] almost defined as a litmus te s t in determ ining
whether race was a predominating factor. So we deter
mined it was necessary in this plan not to have any rural
counties split.
Q. Did you construct the counties in the Eleventh to
comply with one person one vote?
A. Yes, but a more overriding reason was to keep the
164
political subdivisions intact. 1 say tha t because some
times when you tend to focus on keeping whole counties
intact that has the effect of increasing the overall devia
tions of the plan. So I would say that respect for political
subdivisions was the higher criteria.
Q. Were you concerned with retrogression?
A. As it relates to the border between the Fifth and
Eleventh Districts that was a factor. Because the Fifth
is an historical d istric t th a t is pro tected under the
Voting Rights Act. We were concerned tha t we pre
serve the Fifth D istrict to offer the voters in metro
Atlanta an opportunity to elect the candidate of their
choice.
Q. Were you concerned about maintaining contiguity?
A. That is always a factor.
Q. What about protecting incumbents was that a fac
tor?
A. It was not a predominant factor, but it was consid
ered on the several occasions. I didn’t know the home
addresses of the incumbent Congressperson. Usually
when I draw a plan I prefer to have the home addresses
of the incumbents, but th a t was not available on the
State system and that was not taken into account here.
Q. What is the total deviation in this plan?
A. .93.
Q. Would it be up [sic] possible, by splitting some
additional counties in the plan, to lower that deviation?
A. Yes. Generally if you draw a plan at a lower level
of geography you can get more precise and bring the
deviations down to almost zero if you want to.
I have been involved in districting in Alabama where
165
the option that was presented was that the [sic] every
district had to have exactly the same number of people.
Q. In constructing the Eleventh District was race a
predominant factor that you took into account?
A. No.
Q. Did you subordinate traditional State redistricting
principles in the construction of this plan?
A. No. If I can elaborate. I reviewed the State’s his
tory in congressional redistricting all the way back to
the turn of the century to learn how congressional dis
tricts in Georgia were drawn. Then I focused on the
1970 and 1980 plans. I believe over there is the map of
the 1980’s congressional plan. One thing that struck me
is that the State of Georgia has a history of drawing con
gressional districts that are vast in size.
In the 1970’s as well as in the 1980’s congressional dis
tr ic ts encom passed 25 to 30 counties. The N in th
stretched all the way from the South Carolina border all
the [way] down into Gwinnette County.
Another feature that I observed about the congres
sional districts is the State has a tradition of linking
together urban and rural areas in congressional district
ing. What that led me to conclude, on the question of
construction of the congressional districts, was that geo
graphic compactness is not an absolute. In terms rela
tive to the drawing of say a school board district I would
look for a more insular group. But in a state that has
congressional districts that stretch all the way from the
Florida border to Jones County north of Macon, in look
ing at that type of history, I thought it permissible to
draw such a district when you are balancing the rights
of black and white voters.
166
Q. Did you a ttem p t to address the issue of land
bridges and avoid their use in your construction of the
Eleventh?
A. Yes. Specifically in the existing plan there has
been much talk about the so called Henry County land
bridge. We felt it is was necessary to not have such a
land bridge in the Eleventh Congressional District so
that was eliminated.
Q. Could you have put all of Henry County into the
Eleventh or taken it all out?
A. There are many options in the drawing of congres
sional districts because of their vast size. Essentially
you link together a string of rural counties or urban
counties and you come up with a congressional district
with a population of approximately 589,000 people. We
linked together a string of counties in constructing a
number of options for the Eleventh District. Some of the
counties were overwhelmingly white and some rela
tively mixed almost 50 percent black and white and
some were slightly black in terms of majority population
and we joined together those counties to the [sic] make
the Eleventh Congressional District.
Q. Let me move on to the F irst District and ask you
what changes you made in the F irst that are different
from the existing F irst Congressional District?
A. The primary change in the F irst had to do with
the first change I mentioned regarding the Eleventh.
The most unacceptable feature of the Eleventh, from
everyone I heard from, was the extension of the so-
called tail into Effingham and Chatham Counties. Once
that was removed it was necessary to take population
out of the First so that population was taken out of the
First and put into the Eighth. There were four counties
that were moved Montgomery, Tattnal, Toombs—three
167
counties, and those three counties were taken out of
F irst and put into the Eighth.
Q. That’s because Effingham and Chatham Counties
were included in their entirety in the First?
A. Yes. That was done to balance out the population
in terms of one person and one vote. We moved Clinch
County from the Eighth back to the First,
Q. Are there any split counties in your version of the
First?
A. No, all of the counties are intact.
Q. I asked you a questions earlier with regard to the
E leventh about your decision to keep the counties
intact. Would that answer be the same if I ask you why
you kept the counties intact in the First?
A. Yes.
Q. Look at the Second. What changes did you make in
the Second?
A. When I reviewed the Second there were two prin
cipal areas that jumped out at me and a couple addi
tional areas that I was concerned with. The two princi
pal areas were the way in which the Second extended
up into Meriwether County. Our first goal was to correct
that and to take Meriwether County out of the Second
District altogether.
In ACLU1A Meriwether as well as Tolbot Counties
are removed from Second Congressional District.
Q. Why did you remove Tolbot County?
A. Once I made additional changes, which I will speak
to in a minute, the remaining reason was to balance out
the population in term s of one person one vote.
168
The other feature I was concerned about, as it relates
[to] the existing Second Congressional District, was the
way the d istric t en ters Bibb County in the City of
Macon. There was very narrow land bridge, if you will,
in the way the district enters Bibb County. Our goal was
a [sic] correct that and ACLU1A does tha t by taking
additional voters out of Bibb County and the City of
Macon and including them in the Second District. A
number of those voters who were added were white vot
ers. There were some other changes we made to the
Second District.
Q. What were they?
A. There was a string of counties that were split in
the Second District and they shared a border with the
Eighth District. Our goal was to eliminate these county
splits and return the counties to being whole.
Q. Can you tell us what those counties are?
A. Those counties are Lowndes, Colquitt Dougherty,
Lee, Crisp, Dooly, Houston, Bleckley, Twiggs and
Crawford.
Q. Did you reaggregate those counties for the same
reason that you reaggregated counties in the other dis
tricts?
A. Yes.
Q. Did the rural urban factor come into play, in any
way, in the construction of the Second?
A. Yes, in that, as I mentioned earlier in reviewing
the State’s history and particularly with regard to the
construction of Eighth, Seventh and Ninth Districts, we
started to link together urban and rural areas. In doing
so there were a number of urban areas currently in the
169
Second th a t we have sought to re ta in . They are
Columbus, Macon, Valdosta and Albany.
Q. Do you recall how many total unsplit counties are
currently in your version of the district?
A. You have to excuse me. I have been looking at a
large number of plans and after a while you start to for
get how many counties are in each plan. 26 it [sic] what
it looks like.
Q. The only counties that are split[,] that are split in
Bibb and Muscogee in the Second?
A. That’s correct.
* * *
Q. Were you aware of whether or not the county gov
ernments of Bibb and Muscogee had passed resolutions
requesting that those counties remain split in any redis
tricting plan?
A. Yes.
Q. Did you take that into consideration in splitting
those counties is [sic] ACUL1A?
A. That was a factor.
Q. Are the splits in ACLU1A the same as the splits
in the existing plan?
A. No, they are not.
Q. In what ways are they different and why they are
different?
A. I have described the way in which the Second
enters Bibb County and the way Bibb County is split in
the current plan. The split in ACLU1A in some ways
resembles that split. In ACLU1A Bibb County is split
170
between the F irst and Third and in ACLU1A the same
is true for Muscogee County.
Q. Were these splits based, in any way, on race?
A. No, they were not.
Q. What were the other factors?
A. I t was the wish of the county commissioners to
have more than one congressional representative. The
second was the State had historically linked rural and
urban areas together to form congressional districts.
Q. Have we discussed the [sic] all of the changes you
have made to the Second District?
A. Not in detail, but we have covered them.
Q. What, if any, changes were made to the Third dis
trict in ACLU1A compared to the existing plan?
A. The changes to the Third D is tr ic t basically
resu lted from the changes made in the Second and
E leventh D istricts. Tolbot was moved to the Third
D istric t having previously been in the Second and
Baldwin County was included in the Third District. The
other principal change that occurred had to do with
Clayton County.
I have mentioned that we were concerned about hav
ing taken a portion out of DeKalb out of the F ifth
District. To compensate for voters, who were predomi
nantly black, who were taken out of the Fifth District to
make the county line the dividing line between the Fifth
and Eleventh. When those voters were taken out of the
DeKalb C ounty portion of the F if th we basically
extended the Clayton County portion of the F ifth
District. So we extended the Clayton County portion of
the Fifth District. So there were additional voters taken
out of the Third.
171
Q. There are 11 intact counties in the Third; is that
right?
A. Yes.
Q. I won’t ask you to repeat all of your testimony, but
did you keep those counties intact for the same reasons
you did in the other districts?
A. Yes.
Q. As to the Third District, as a whole, was race a
predominanting [sic] factor in the construction of the
plan?
A. No, race was not a predominanting [sic] factor in
the construction of the Third District. Nor was it a pre
dominant factor in the construction of this plan.
* * *
Q. Will you explain what, if any, changes were made
to the Fourth District and tell us why they were made?
A. The changes in the Fourth District were minimal.
P rim arily it involved some p rec in c t changes in
G w innette County betw een the F o u rth and Tenth
Districts and they were made to compensate for shifts
in population that occurred in the Tenth once we cor
rected that constitutional defect in Richmond County
and put Wilkes County back as a whole country.
Q. Was the purpose for those changes predominantly
based on race?
A. No. The purpose was predominantly one person
one vote.
Q. What changes did you make in the Fifth and why?
A. There was an additional change in the Fourth,
which was in DeKalb County in the southern most part
172
in the City of Decatur there were a couple of precinct
chages. That was done to balance out the population.
Q. What changes were made in Fifth and why?
A. The most striking change was to eliminate this
extension of the Fifth into DeKalb County and to make
th e DeKalb F u lto n C ounty line the d iv id ing line
between the two districts and that took some population
out of the Fifth which was compensated for with the
extension into Clayton County.
In addition to th a t the northern part of the F ifth
District was enhanced by removing some of the border
to [sic] in the most northern part of Fulton County.
Q. Was the predominant purpose of those changes
based on race in those counties?
A. No, those chages were made to balance out popu
lation. A very important goal was compliance with the
Voting Rights Act to make sure that the redesigning of
the Fifth District did offer African America[n] voters a
full opportunity to elect the candidate of their choice.
Q. What was the racial composition of the changes
that were made? In this area of DeKalb, for example,
was that predominantly black?
A. Yes, it was. The area of DeKalb that was removed
from the Fifth that portion of Atlanta on both sides of
the county line the population is pretty solidly black.
Q. And the area of Clayton County that you put in
was it substantially of the same racial composition of the
area that you took out?
A. A substantial portion of the popualtion was black,
but there were also some whites who tended to vote
Democratic.
173
Q. So was it a non-racial change or did you exclude
more blacks than you brought in by making the DeKalb,
Clayton County exchange?
A. The portion of the Fifth District that was taken
out of DeKalb County to make the dividing line the
county line that portion was, I would say, about 95 per
cent black. The portion of Clayton County tha t was
included in the Fifth was not as heavily black as it was
in the DeKalb County portion.
Q. What was the reason or purpose for making that
exchange?
A. To prevent retrogression.
Q. Look at the Sixth Congressional District. What
changes did you make there?
A. The changes there were very minor. There were a
couple of precinct chanes in Gwinnette County similar to
what I have described in the Fourth and the Tenth. I
believe th e re w ere two p recinct changes in Cobb
County in the area where the Seventh is next to the
Sixth.
Q. So there were no counties added or taken out?
A. That’s right.
Q. W hat was the purpose of the changes in the
precinct?
A. That was to comply with one person one vote. In
addition this option was an attempt to the [sic] correct
th e constitu tional defect in the E lev en th and the
assumed constitutional defect in the Second and to pre
serve the remaining districts, as much possible, intact.
Q. What about the Seventh, what changes were made
there?
174
A. The principal change would be the inclusion of
Meriwether County into the Seventh and that was done
for the purpose one person one vote after Meriwether
was removed from the Second because it had been
determined by the panel of judges as being unaccept
able in terms of its appearance.
Q. No other counties were added or deleted?
A. No.
Q. Cobb is split in the Seventh in your version?
A. Yes, in pretty much the same way it was.
Q. Why did you retain that split?
A. My instructions from you in this option was to pre
sent to the Court a plan which corrected the constitu
tional defect in the Eleventh and the assumed constitu
tional defect in the Second and to keep the remaining
districts unchanged as much possible.
Q. Had you corrected the split in Cobb would be [sic]
you have had to make other changes to compensate for
the addition or loss of population?
A. Yes, sir.
Q. Would that have entailed significant changes?
A. Cobb County is part of metro Atlanta so the popu
lation there, like in Fulton, DeKalb and Gwinnette, is
very dense. So when you move a precinct in Cobb
County it is like moving a rural county.
Q. W hat changes did you m ake in the E ig h th
District?
A. In the Eighth district the changes were primarily
made to compensate for the changes I described in the
Second and the changes that were made in the F irst and
175
the changes th a t were made iri the E leventh . So I
included the counties Montgomery, Toombs and Tattnall
into the Eighth District and they were previously in the
First. I excluded Clinch County as I mentioned previ
ously and reunited Houston as one county and then the
extension into Bibb County does not exist.
Q. Now, you excluded Clinch was that to comply with
one person one vote?
A. Yes.
Q. Were the changes you made in the Eighth based
on race?
A. No. One person one vote was the major factor. The
other predominant factor, in terms of the changes in the
E ig h th , was to keep in ta c t political subdivisions.
Because the string of counties in the Eighth that shared
their border with the Second, as mentioned earlier there
were a number of counties that were previously split in
the current plan, all of those counties were reunited.
Q. Did you make any changes in the Ninth District?
A. No, there were no changes in the Ninth.
Q. So the Ninth remains as it currently is?
A. That’s correct. It stays as it currently is.
Q. The last district is the Tenth what changes did you
make there?
A. Wilkes County is a whole county in the Tenth
resulting from eliminating this split in the Eleventh
District. In addition in Richmond County I took in a
larger portion of the white population in Richmond
County and moved it from the Tenth and added it to the
Eleventh. Those were the principal changes. As I men
tioned earlie r th e re were also a couple of precinct
changes in Gwinnette County.
176
Q. W hat was the purpose of those changes in the
Tenth?
A. It was to balance out population.
Q. Were any of the changes made in the Tenth based
on race?
A. No.
Q. Is there any single rural county tha t is split in
ACLU1A?
A. No. In drawing ACLU1A we were very particular
in reuniting all of the rural counties which are currently
split in the existing congressional districting plan and
which were split in some of the earlier versions of plans
we had developed. That was done primarily because
there was much talk about the splitting of rural counties
as being sort of a litmus test and we have eliminated all
rural county splits. The only counties that are split are
in this plan are the larger more densely populated met
ropolitan areas of the State.
Q. Would it be possible to draw other plans tha t
were in the so called least disruptive plan category or is
this the only version you could possibly do?
A. Could you restate your question?
Q. Is it possible to draw other plans in the so called
category of least disruptive or least change, but cure the
constitutional defects in the existing plan or is this the
only possible plan?
A. Yes, sir, I believe it is possible to draw other plans.
I have p repared an option which leaves Richmond
County completely out of the Eleventh District. So it
depends on what your motivation is.
In the Legislature there was a feeling that some peo-
177
pie would not pass a plan if Richmond County were split
or was part of the Eleventh. Depending on your goal it’s
possible to prepare other options.
Part of my responsibility in working with the special
session was to review the plans of others for fairness
and I reviewed the Amicus Plan and the version before
the one called Amicus Plan. That plan came pretty close
to what I would call the least change plan, but the
Eleventh District was not in Richmond County.
Q. Did Mr. Hill prepare this plan ACLU1 A?
A. No, I supervised its preparation and I was sur-
prise[d] when I heard that here today. He did not pre
pare that plan.
Q. I’m not going to ask you in detail about the other
plans, but I think it would be helpful for you to tell us
briefly what you attem pted to do when you prepared
the first three plans that were filed with your orininal
declaration and the supplemental declaration that were
received in evidence.
A. Those plans were developed either shortly before
the special session convened or during the special ses
sion. The three plans you mentioned were developed
with the express purpose of presenting options before
the legislative black caucus. The goal was to develop a
plan that could achieve the number of votes necessary
to pass both houses. In preparing those plans the over
riding goal was to show that it was possible to be fair to
both black and white voters in the s[t]ate and to comply
with the Court’s directive in Miller v Johnson.
In that regard the first plan had some county splits
that were eliminated and the goal was eliminating those
county splits. As we moved from the first to the third
plan our goal was to remove all of the splits in the rural
178
counties and to also develop a plan that would basically
serve as a basis for the formation of a coalition between
black and white Democrats and be able to pass both
houses.
By the time we got to the third plan our goal was to
eliminate all splits in rural counties and to redraw all of
areas of [sic] in the state outside of the Fifth, Eleventh
and the Second pretty much the way white Democrats
wanted them drawn with the hope that it would allow
white and black Democrats to form a coalition and get a
plan passed by both houses.
For example in tha t plan the Seventh District was
drawn exactly as the Speaker wanted it. The F irst is
draw[n] exactly the way the Chairman of the House
Reapportionm ent Com mittee w anted it. The Ninth
District was d[r]awn close to the way the Lieutenant
Governor wanted it drawn.
Q. Was that plan endorsed by either house?
A. The plan which passed in the House was derived
from that plan.
Q. Which plan?
A. The third one. The plan which passed the House
was derived from the plan that I just described and the
principal difference was in the Eleventh District. The
Eleventh District did not encompass all of the rural
counties tha t that plan encompassed. We stopped in
Washington County and did not go all the way in. Other
than that the plan tha t passed in the House is almost
identical to the plan I described here.
MR. McDONALD: Thank you, very much.
179
CROSS-EXAMINATION
BY MR. PARKS:
Q. Mr. Carter, with respect to the ACLU1A plan is it
your position that the configuration of the Eleventh
District in that plan is mandated by the Voting Rights
Act?
A. My position is th a t ACLU1A was draw n in
response to a request I received from Mr. McDonald to
prepare a least change plan.
Q. I understand that, but is it your position that in
the racial consideration that you testified to, leaving
aside whether is was a predominant, intermediate or
insignificant factor, the racial considerations do you feel
they were mandated by the Voting Rights Act or are
just a product of your attempt to draw a minimally dis
ruptive plan?
A. I believe the word mandated is one Pm not sure of.
Q. Required?
A. I will answer that in the way I understand it Pm
not sure it’s mandated. There are lots of differences of
opinion in the aftermath of Miller v Johnson and Shaw v
Reno as to how you balance the interest of the require
ments of the Voting Rights Act with what is now the law.
* * * * *
180
MR. PARKS: I have not heard an answer.
A. Essentially I believe the jurisdiction between state
governments and voting right advocates are confused on
that question. I sat through a number of conferences. I
sat through the special session of the Legislature. I have
talked to a number of people and find that there is a gen
eral disagreement as to how you balance the interests of
the Voting Rights Act against the descision [sic] of this
Court and the Supreme Court in Johnson v. Miller. I do
believe as a result of Thronberg v. Jingles a state has an
obligation to draw a district that is fair to minority voters
if the geographical distribution of voters is such that such
a district can be drawn. I believe the Eleventh District is
such a district that can be drawn because I believe in the
past the state has drawn districts just as vast. Does that
answer your question?
Q. I don’t know.
A. I t’s more complicated than that. I don’t know.
Q. But you can’t say for sure whether it is or not?
A. I have answered your question, basically in the
way I just did.
Q. I am not suggesting there is a definitive answer,
but I didn’t hear a definitive answer.
A. I will tell you what it depends on is geographic
compactness. Geographic compactness is a very subjec
tive term. What is geographically compact to me may
not be geographically compact to you. That’s why that
question is so difficult to answer.
Q. Now, if you have the same difficulty as you had
181
with in the question I asked you about the ACLU mini
mal disruption plan you can just say I have the same dif
ficulty without going into that long response.
In your opinion is th e E lev en th C ongressional
District, as configured in this Amicus R Plan, required
by the Voting Rights Act?
A. The same difficulty.
Q. Mr. Carter, you told any number of people during
that session that to be politically acceptable to the black
caucus there had be a second congressional district that
was majority black?
A. Yes, sir.
Q. Regardless of w hether it was required by the
Voting Rights Act or not for the black caucus to support
any plan it had to include a second majority black dis
trict?
A. I do not recall making any reference to the Voting
Rights Act.
Q. But in you own mind that was the reality?
A. Yes.
Q. That’s why I asked you the question because my
reading of the position of the Abrams Intervenors, as
you have described it, is there isn’t a single plan ACLU,
black caucus or your organization th a t you can say
under oath is required by the Voting R ights Act in
terms of a second minority, majority district?
A. I believe that we are in gray area of the law. It
used to be clear a couple of years ago and it is not as
clear now as it used to be.
182
* * * * *
Q. But you would agree that there could still be clear
cases wherein a minority district would be required by
the Voting Rights Act?
A. I believe that’s true.
rfc sf:
Q. You don’t make the distinction tha t the district
was not reflective of a true community of interest. You
are saying it isn’t as bad off as some others in the past
and therefore why not?
A. Repeat that, please.
Q. You are not saying the Eleventh District is reflec
tive of any real true community of interest. Instead you
are saying it has a sufficient degree of geographic com
pactness to merit consideration?
A. I haven’t done an analysis for this presentation
here today.
* * * * *
REDIRECT EXAMINATION
By MR. MCDONALD:
Q. Mr. Carter, let me direct your attention to the 1972
Congressional red is tr ic tin g plan to Congressional
District One?
A. Yes.
Q. Can you tell us whether in this district you have
any counties that strung together in a single file march
ing from the south up to the north?
A. Yes, that was one of the districts that struck me
when I was struggling with this plan in terms of draw-
183
ing it. One of the first things that struck me was the
S tate’s history in drawing these type districts. There
you have a district that runs from the Florida border up
to Augusta.
Q. Is that a land bridge?
A. The te rm land bridge is not one th a t I have
embraced. I have never use that on my own.
Q. Let me hand you this map and ask you to identify
it.
MR. McDONALD: I would like to [have] that marked
as Abrams Intervenors Exhibit 37.
Q. H anding you w hat has now been m arked as
Abrams Intervenors number 37, can you identify that
for me?
A. It looks like the plan which passed in the Georgia
House.
Q. Judge E denfield asked th e question e a rlie r
whether or not it is possible to construct a plan that con
tained two minority majority districts that didn’t have a
wall across the state. Can you tell me whether or not
that is the case with Abrams Intervenors number 37?
A. Yes, sir. You can see it again in the Eleventh dis
trict which does not stretch across the state. It extends
from DeKalb to Jefferson County. So there are still a
couple of counties between the Eleventh District and
the state border. The question of minority majority is
really tricky. If you look at black voting age population
it is only 50.41 percent. It is a very integrated district.
Q. What is the total minority population in this dis
trict?
A. The district is 53.37 percent black.
184
N. ARGUMENT: DAVID WALBERT
[Excerpts from Hearing Testimony, 10/30/95-10/31/95 Tr.
p. 378, lines 18-25; p. 379, lines 9-13 and 21-23; p. 380,
lines 2-24; p. 384, lines 2-4.
Section Five doesn’t apply to this Court. If this Court
draws a plan Section Five doesn’t put a bench mark on
this Court and this panel of three judges. There was a
Section Five discussion earlier with Mr. Brumby about
w hether if the Speaker subm itted a plan is th a t a
Section Five covered plan. He has authority in his brief
suggesting that may be.
* * *
Be that as it may the bench mark for Section Five if
this were a legislative plan, and I think the Court should
put itself in the shoes of the legislature when it is draw
ing a plan, the bench mark here is clearly the ‘82 plan.
* * *
It is our position that the ‘82 plan, being the last con
stitutional plan, has to be the bench mark.
MR. WALBERT: Yes, your Honor, that’s definitely
our position.
To address the Section Two question. It is pretty clear
th a t this Court in devising a rem edy has to make a
determination as [to] whether Section Two does or does
not apply to any plan that is adopted by this Court. We
have a pretty strong opinion on this. I guess the first
thing I will say is we take very strong issue with the
Justice D epartm ent’s view th a t they advocate th a t
Section Two is required. I will say this with all due
respect to the Justice Department lawyers, but what
they are doing here is exactly what they were doing in
1991. The old game, if you will, that this Court criticized
185
and which the Supreme Court criticized, the old game
was if you didn’t do it disprove why you didn’t do it and
since you can’t disprove why you didn’t do it intentional
and purposeful were applied. The old game was intent
and purposeful were applied because of your inability to
disprove that you didn’t go to Chatham because of race
or didn’t go to in Richmond County because of race. It
all boiled down to an inability to disprove.
Section Two we very much disagree that Section Two
mandates there be two black members majority dis
tricts under its mandates.
186
0. TESTIMONY OF LINDA MEGGERS
[Excerpts from Hearing Testimony, 10/30/95-10/31/95
Tr. p. 430, line 3-p. 434, line 8, p. 438, lines 12-17]
DIRECT EXAMINATION
BY MR. PARKS:
Q. Were you involved during the special session with
regard to attempting to redistrict the State of Georgia
to correct the unconstitutionality that had been found
within the existing redistricting plan?
A. Yes, sir.
Q. During the course of that involvement did you
make an effort to construct a second minority majority
district that was consistent with your reading of Miller
versus Johnson and the Voting Rights Act?
A. Yes, sir. I started out from two different angles.
What I did is I sat down by myself after the decision
came out and before anyone could talk to me and I read
it three times to see how I understood it. I know I’m not
an attorney and I can’t quote the law of the case, but I
feel I have to understand it before I can help the people
I work with. I t seemed to me there was a lot of talking
about the only thing that was unconstitutional about the
E leventh D istrict was the fact it had an arm going
through Screven and down to Chatham and that’s all
we needed to do to fix it. As I looked at it and as I
understood the Court decision from both this Court and
the Supreme Court the idea that I think came out of the
Supreme Court decision was why did we have [to] draw
the Eleventh District like we did in 1991?
To be honest with you I was really partly at fault for
drawing it like we did in 1991. In my understanding of it
I thought the Department of Justice was going to make
187
us do. P rior to the work in 1991 there had been an
understanding arrived at by the leadership in the legis
lature with the black caucus that we are going to draw a
second minority district and you can have it. I think at
that time we understood that to mean under our normal
interpretations of the Voting Rights Act meaning com
pactness, contiguous and community of interest.
I looked at that in two different ways to draw that.
One of the major efforts I made by myself a t that time
was to look over to east central Georgia. If you will look
at the race map there is a heavy concentration up here
and I think we talked about how all that got added to
DeKalb county to begin with. I looked into the metro
area and none of [the] ways I drew ever showed a total
of a 50 percent minority population. At tha t time we
were talking about an effective black district being 60
percent. At the first public hearing we were presented
by the head of the Republican party a map that literally,
he said, had been faxed to him about an hour before
tha t. That plan drew it from south DeKalb over to
Augusta and then to Macon. The initial response in our
office was that’s ridiculous. We don’t have draw that do
we? I was the one who said if it was there we have to
draw it. It was said that it doesn’t make any sense and I
said maybe not, but I think we may get in trouble with
the Justice Department if we don’t draw the State of
Georgia like that and I think that was probably the main
reason they went in that direction.
After this decision in Miller v Johnson came out I sat
down and I read that again and I said why would draw
it that way? If you are going to look at motivation what
other reason would you ever have for coming out of
south DeKalb like that? What other reason would you
have for putting Burke County with south DeKalb?
What would Johnson County ever have in common with
188
south DeKalb? So even looking a t cleaning up the
Eleventh from the way we drew it in 1991 I didn’t think
satisfied the constitutional requirements. I needed to
look at another way to create a second minority major
ity district.
All of us were aware that Speaker Murphy, even after
the court decision, said that he thought you ought to
have two minority majority seats in Georgia. If I look at
the guidelines and I hear people say the Supreme Court
left all of this confusion. I don’t think they left that much
confusion in my mind. They told me that if black popula
tions were compact, contiguous and had a community of
interest they should be kept together. I have always
understood that and that didn’t bother me. I thought
th a t was a trad itional in te rp re ta tio n of the Voting
Rights Act. They also said th a t you use traditional
redistricting principles. I had been working on this since
1971 and I thought I had a pretty good handle on what
traditional redistricting principles were in Georgia.
I looked at that and I looked at the race map because
I do think that you take race into consideration. They
are 27 percent of the population and if we are going to
look at distributing political power which I think that’s
what redistricting is all about. The concentrations that
you have are still in the metro Atlanta area. That is the
one concentration you have if you are looking at com
pactness, contiguity of population and community of
interest, you have urban areas, you have areas of inter
est. Beyond the color of skin I think there is probably
other interests that blacks have in common, but putting
that aside there should be areas of interests. That was
the place where I started, to be honest with you.
* * *
189
Q. Consistent with your understanding of the deci
sion in Miller v Johnson and the Voting Rights Act a
second minority majority district was not, at least in
your opinion, possible?
A. Not one that said 50 percent on the page in 1990,
no, sir, it was not.
1995 House-Passed Plan
(MSLSS)
Library; Georgia
Plan: !it16620124hse
□
■
0
a
a
EB *>
Gorily (fcxnrtarios
I
If
|
a
June 19,1996
User: Temporary Account
Room: 941
Plan: LIT16620124HSE
191
Date: 06/20/96
Time: 9:32 AM
Report: housepassed. rep
Population Summary Report Plan MSLSS
(House Passed)
D ist TO TA L D E V IA T IO N D E V % W H IT E % B L A C K %
1 588541 -388 -0.07 67.58 30.55
2 587583 -1346 -0.23 59.54 39.21
3 589276 347 0.06 73.09 24.80
4 589277 348 0.06 55.02 40.36
5 590666 1737 0.29 39.57 58.01
6 589268 339 0.06 91.36 5.80
7 588249 -500 -0.08 85.15 13.70
8 587912 -1017 -0.17 67.92 31.07
9 589420 491 0.08 94.68 3.65
10 588495 -434 -0.07 85.48 12.69
11 589349 420 0.07 61.78 36.67
T otal P o p u la tio n : 6478216
Id e a l P o p u la tio n : 588929
M ean D ev ia tio n is : 0
M ean P e rc e n t D ev ia tio n is : 0
L a r g e s t P o s itiv e D ev ia tio n is : 1737
L a r g e s t N e g a tiv e D ev ia tio n is : -1346
O v era ll R an g e in D ev ia tio n is : 3083
O v era ll R an g e in D ev ia tion% is : 0.52
I
6
1
Lewis/Gingrich Plan
(AMICUSR)
Library: Georgia
Plan: Iit16620124am icusr
OtSIWGtS
a >
m
a 3
■ -
s >
□ «
a '
s »
■ »
ED 10
Prepared by:
United Stales Department o f Justice
Washington, D.C., 20530
June 25,1996
193
Date: 06/25/96
Time: 9:30 AM
Report: popsum. rep
User: Temporary Account
Room: 941
Plan: LIT16620124AMICUSR
Population Summary Report Plan AMICUSR
D is t TO TA L D E V IA T IO N D E V % W H IT E % B L A C K %
1 589329 400 0.07 67.62 30.49
2 591188 2259 0.38 49.09 49.41
3 588137 -792 -0.13 78.74 20.03
4 583497 -5432 -0.92 85.03 9.72
5 586998 -1931 -0.33 40.01 57.33
6 592594 3665 0.62 91.54 6.03
7 594440 5511 0.94 87.06 11.81
8 584623 -4306 -0.73 72.07 26.74
9 590603 1674 0.28 93.56 4.76
10 589636 707 0.12 72.04 25.86
11 587171 -1758
oCO©1 44.05 54.65
Total P o p u la tio n : 6478216
Id e a l P o p u la tio n : 588929
M ean D ev ia tio n is : 0
M ean P e rc e n t D ev ia tio n is : 0
L a rg e s t P o s itiv e D ev ia tio n Is : 5511
L a rg e s t N e g a tiv e D ev ia tio n is : -5432
O verall R an g e in D ev ia tio n is : 10943
O verall R an g e in D ev ia tion% is : 1.86
Abrams Plan A
(SRC/ACLU Plan A)
Library: Georgia
Plan: iit16620124abram sa
nisi facts
a ’
a
a *
m -
m *
□ ■
a i
s »
■ »
a »
» «
Cospty ikxrxlaiioe
Prepared by:
United States Department o f Justice
Washington. D C ,, 2 (6 3 0 '
June 2 2 ,1996
195
User: Temporary Account
Room: 941
Plan: LIT16620124ABRAMSA
Date: 06/22/96
Time: 4:29 PM
Report: popsum. rep
D ist
Population Summary Report Abrams Plan A
TO TA L D E V IA T IO N D E V % W H IT E % B L A C K %
1 588455 -474 -0.08 67.74 30.56
2 588499 -430 -0.07 44.11 54.45
3 588995 66 0.01 77.39 21.09
4 590178 1249 0.21 86.38 8.67
5 588878 -51 -0.01 40.84 56.70
6 588851 -78 -0.01 89.34 7.80
7 588720 -209 -0.04 85.96 13.07
8 588846 -83 -0.01 73.98 24.81
9 589360 431 0.07 96.38 2.48
10 588869 -60 -0.01 77.25 20.49
11 588565 -364 -0.06 41.65 56.56
T otal P o p u la tio n 6478216
Id e a l P o p u la tio n 588929
M ean D ev ia tio n is 0
M ean P e rc e n t D ev ia tio n is 0
L a rg e s t P o s itiv e D ev ia tio n is 1249
L a rg e s t N e g a tiv e D ev ia tio n is -474
O verall R an g e in D ev ia tio n is 1723
O verall R an g e in D ev ia tio n 'll is 0.29
Abrams Plan C
(SRC/ACLU Plan C)
Library: Georgia
Plan: Iit16620124abram sc
DISIHSCIS
June 19, i 996
PI
LE
: t
o!
»n
.»
i*
Su
ti»
nt
iy
d»
t*
7m
ie
e2
Q
l2
it«
fe
r«
m
*o
/*
b?
,c
197
User: Temporary Account Date: 06/20/96
Room: 941 Time: 8:54 AM
Plan: LIT16620124ABRAMSC Report: abramsc2. rep
Population Summary Report ABRAMS Plan C
D is t : TO TA L D E V IA T IO N D E V % W H IT E % B L A C K %
1 586191 -2737 -0.46 68.73 29.57
2 590278 1350 0.23 48.71 49.80
3 589649 721 0.12 76.60 22.06
4 590029 1101 0.19 85.20 9.49
5 590261 1333 0.23 40.29 57.09
6 590284 1356 0.23 88.42 9.44
7 584478 -4450 -0.76 91.24 7.75
8 587761 -1167 -0.20 70.94 27.68
9 588747 -181 -0.03 90.41 7.69
10 590248 1320 0.22 75.96 22.10
11 590290 1362 0.23 44.86 53.66
Total P o p u la tio n : 6478216
Id e a l P o p u la tio n : 588929
M ean D ev ia tio n is : 1
M ean P e rc e n t D ev ia tio n is : 0
L a rg e s t P o sitiv e D ev ia tio n is : 1362
L a rg e s t N e g a tiv e D ev ia tio n is : -4450
O verall R an g e in D ev ia tio n is : 5812
O verall R an g e in D ev ia tion% is : 0.99
Abrams Plan ACLU1A Library: Georgia
Pian: Iit16620l24aciu1a
Prepared by:
United Stales Department o f Justice
Washington, D.C., 20530
June 22, J 996
i
?-
I
I
5
199
User: Temporary Account
Room: 941
Plan: LIT16620124ACLUIA
Date: 06/23/96
Time: 12:35 PM
Report: popsum. rep
Population Summary Report Abrams Plan ACLU1A
D ist T O T A L D E V IA T IO N D E V % W H IT E % B L A C K %
1 590529 1600 0.27 68.51 29.79
2 589281 352 0.06 48.74 49.75
3 590244 1315 0.22 78.26 20.11
4 585946 -2983 -0.51 84.80 10.19
5 590241 1312 0.22 38.60 58.98
6 589187 208 0.04 91.36 5.97
7 590462 1533 0.26 84.53 14.36
8 591442 2513 0.43 71.06 27.58
9 586222 -2707 -0.46 94.65 3.67
10 586262 -2667 -0.45 78.69 19.66
11 588450 -479 -0.08 42.14 56.27
Total P op u la tio n : 6478216
Id e a l P o pu la tion : 588929
M ean D ev ia tio n is : 0
M ean P e rc e n t D ev ia tion is : 0
L a r g e s t P o sitiv e D ev ia tion is : 2513
L a r g e s t N e g a tiv e D ev ia tion is : -2983
O vera ll R an g e in D ev ia tion is : 5496
O vera ll R an g e in D eviation% is : 0.94
Plaintiffs' Remedy 4 Library: Georgia
Pian: Iit16620124rem edy4
■ •
S3 J
0 »
□ »
B y
m
C anty iiandAhott
Sc8ie in Milos
0 S.S 13 26 39
Prepared by:
United Stales Department o f Justice
Washington, D.C., 20530
June 22,1996
I&9
201
Date: 06/20/96
Time: 9:41 AM
Report: popsum. rep
User: Temporary Account
Room: 941
Plan: LIT16620124REMEDY4
Population Summary Report Plaintiffs’ Remedy 4
D i s t : TO TA L D E V IA T IO N D E V % W H IT E % B L A C K %
1 : 588541 -388 -0.07 67.58 30.55
2 : 594107 5178 0.88 59.29 39.47
3 : 589774 845 0.14 73.61 24.18
4 : 589277 348 0.06 55.02 40.36
5 : 583644 -5285 -0.90 39.03 58.62
6 : 587419 -1510 -0.26 91.34 5.81
7 : 593007 4078 0.69 85.24 13.61
8 : 587912 -1017 -0.17 67.92 31.07
9 : 589420 491 0.08 94.68 3.65
10 : 584570 -4359 -0.74 85.57 12.60
11 : 590545 1616 0.27 61.69 36.77
Total P o p u la tio n : 6478216
Id e a l P o p u la tio n : 588929
M ean D ev ia tio n is : 0
M ean P e rc e n t D ev ia tio n is : 0
L a rg e s t P o s itiv e D ev ia tio n is : 5178
L a rg e s t N e g a tiv e D ev ia tio n is : -5285
O verall R an g e in D ev ia tio n is : 10463
O verall R an g e in D ev ia tion% is : 1.78
Plaintiffs' Remedy 4X Library: Georgia
Plan: Iit16 6 20 124new 4x
a <
m >
a 3
□ •
B '
B 9
■ 9
m »
Grxrrty BoivUnos
Prepared by:
United States Department o f Justice
Washington, D.C., 20530
June 19,1996
Ss*
i
t
193
1
203
User: Temporary Account
Room: 941
Plan: LIT16620124NEW4X
Date: 06/20/96
Time: 9:41 AM
Report: Remedy 4X. rep
Population Summary Report Plaintiffs’ Remedy 4X
D ist TO TA L D E V IA T IO N D E V % W H IT E % B L A C K %
1 588541 -388 -0.07 67.58 30.55
2 587583 -1346 -0.23 59.54 39.21
3 589276 347 0.06 73.09 24.80
4 589277 348 0.06 55.02 40.36
5 590666 1737 0.29 39.57 58.01
6 589268 339 0.06 91.36 5.80
7 588429 -500 -0.08 85.15 13.70
8 587912 1017 -0.17 67.92 31.07
9 589420 491 0.08 94.68 3.65
10 588495 -434 -0.07 85.48 12.69
11 589349 420 0.07 61.78 36.67
T otal P o p u la tio n 6478216
Id e a l P o p u la tio n 588929
M ean D ev ia tio n is 0
M ean P e rc e n t D ev ia tio n is 0
L a rg e s t P o s itiv e D ev ia tio n is 1737
L a rg e s t N e g a tiv e D ev ia tio n is -1346
O verall R an g e in D ev ia tio n is 3083
O verall R an g e in D ev ia tion% is 0.52
ROCKDALE
% Black
i— ] < 3 5
H 3 5 - 4 9
m 5 0 - 5 9
□ > = 6 0
Client : REPCAUCUS
Plan : AMICUSR
Type : Congressional
Q < 35
m 35-49
E l 50-59
□ >= 60
% Black
Client : REPCAUCUS
Plan : AMICUSR
Type : Congressional
207
D ISTRICT 11 FROM F IR ST CONGRESSIONAL
P LA N PASSED B Y GEORGIA LEGISLATURE
IN 1991 (S.B. 2EX)
(N O T E : V A P = V oting A ge P opu la tion )
D is tr ic t Total Pop B lack Pop VAP B lack VAP
N u m b e r % D ev ia tio n % o f Total % o f Total % of V A P
REM AIN D ER 158263 70810 111964 45968
OF DISTRICT -73.13 44.74 70.75 41.06
DEKALB 230592 163176 163684 109727
PORTION -60.85 70.76 70.98 67.04
RICHMOND 114910 66539 81691 43744
PORTION -80.49 57.91 71.09 53.55
BIBB 79588 53179 57903 35635
PORTION -86.49 66.82 72.75 61.54
T otals 583353 353704 415242 235074
N u m b e r o f D is tr ic ts : 11
M em b ers P e r D is tr ic t : 1
Id e a l D is tr ic t Size : 588929
A verage D ev ia tio n (% ): 75.24
D ev ia tio n R an g e (%) : -86.49 to 0.00
O verall D ev ia tio n (%) : 86.49
(DATA S O U R C E : 1990 U S C en su s PL94-171 P op u la tio n C o u n ts )
( N O T E : D is tr ic ts n u m b ere d > 200 a re u sed as spec ia l )
( accum ula to rs . T h ey a re n o t included in av g )
( o r % ra n g e calcu lations. )
208
Jgmpreme C ourt of the ptmiefr J^tates
No. 95-1460
U nited States, appellant
v.
Davida J ohnson, et al.
APPEAL from the United States District Court for
the Southern District of Georgia.
The statement of jurisdiction in this case having been
subm itted and considered by the Court, in this case
probable jurisdiction is noted. This case is consolidated
with No. 95-1425 - Lucious Abrams, Jr., G. L. Avery,
Williams Gary Chambers, Sr., and Karen Watson and a
total of one hour is allotted for oral argument.
May 20,1996
209
Jgmpreme (Hourt of the ^fiuteh ^States
No. 95-1425
Lucious A brams, J r ., G. L. Avery
W illiam Gary C hambers, Sr ., and Karen Watson,
APPELLANTS
V.
D avid a J o h n so n , e t a l .
APPEAL from the United States District Court lor
the Southern District of Georgia.
The statement of jurisdiction in the case having been
subm itted and considered by the Court, in this case
probable jurisdiction is noted. This case is consolidated
with No. 95-1460 - United States v. Davida Johnson, et.
al. and a total of one hour is allotted for oral argument.
May 20,1996