Syllabus; Court Opinions

    (SIip Opinion)

NOT& Wh.r. it i! fcolibL. s sylLbur (h.rd!ot ) sill br relrrc4 rr ir
bdnc dorr ia coru(tioo with thir crI. sr thG tiEa tba ooiaiu ir irarcd.
frrivltrhu cou*iturea D D.rt of th!odnboof thr Cqut tir br bou>
il!.dbv th. RemttG! of lEcid.Er for ilc cowcabu of &. rudc Sa*
b,;trrlStata v.'Dttwit L1ljtnbt Co.,80 U. S. 381, 3Sr.

ST]PREME COUBT OE THE T]NIIED,STffTES

Syllsbu$

ITIORNBURG, ATTORNEY GENERAL OF NORTE
CAROLINA, ET ar- ?r. GINGLES sE ar-

APPEAL FROM TEE UNTTED SITATESI DISIRIqT COIIRT FORTEE
EASTEIB}T DTSTTRI T OT' NORTE CAROLTNA

No. 89-1968. Argued December4, 1986-Deeided June 30, 1986

ID $84, the Ncth Camline General Aseembly eoacted e legislativa ne.
ditricting plaa for the Stdslc S€nsts and Eouso of, Beprrsentatives.
Appellees, bleck citizens of North Carolirs who eneregista-d to votn,
hought suit in Federsl Distrist Corui, "halleng'.g ons siugle-aember

redictricdngplan inpdred bleck citizeuC ability to elect repreeentativer
of theb &oice in violation of ! 2 of the Voting Bights Act of 1966.. Aft€r
appelleec brought - rit, but before Eial, $Zwas aueude4 largely in re-

r sP{ltE€ tO l' 'oldcn, 4& V. S. 56, to nqks.clearthst a violatioa

: of E" t by showing: Ascinirstary efiect alone, mthen
aiscrirainatory purposa, and to ectablish as the

i fha 'taarlfq tni ' Saafian 9/ol oc anonia{

:al subdivison from imposing auy voting'qruli-

4
r+
th \/eori \emi\z
thrt *- -tunity than othermernberl of the elec-
tontr $ -- political procese and to elecf repr.eaentatives
of ther ' -,.r ihat the e:rteat to which members of a protected
al'qa ha -"en eleeted to o6ce is one circlrmsEnc€ that mey be consid-
ered. The DisEist Court applied the "totality of the cirstmstbnces" test
se0 forth in $ 2(b) and held that the redistricting plan violatcd $ 2(a) be
canse it resulted in the dilution of black citized votes in all of t}le dis"
puted districts. Appellants, the Attorney Genenl of North Caroline

I



THORNBURG u GINGLES

rt

Syllabus

and otherr, took e dlreet appeal to thle Court wlth respeet to flve ofthe
multlmember dlstrlets.

Hekl: \\eJudgment ls afflrmed ln part and reversed ln part.
690 F. Supp. 846, elfirmed ln part rnd reversed ln part.

Justrcs Bnrrnmr dellvered the oplnlon of the Court wlth reepect to
Perts l, II, III-A, III-8, lV-A, rnilV, eoneludlng that:

l. l{lnorlty voterr who contend that the multlmember form of dlatrtct-
lng vloletes ! 2 must pmve that the uae of a multlmember eleetorar atrue-
ture opereter to mlnlmlze or eaneel out thelr ablllty to eteet thelr pre-
ferred eandida0ea. l4rhlle meny or all ofthe faetorcllsted ln the Senate
Report rnay be relevant to r elalm of vote dllutlon througt aubmergenee
ln multlmember dletrletr, unlesd there ls d eoqfunetlon of the foltowlrrg
clre,umstsnees, the uee of multlmember dlstrlcta generally wlll not lm-
pede the. ablllty of mlnorlty votera to elect representetlves of thelr
eholee. Stated srteclnetly, a bloc votlng mqforlty murt uiwily be abte
0o defeet candldateg rupported by a polltlcally eoheslve, geographleelly
lnsuler mlnorlty group. The felevanee ol the exlatence of raelal bloc
votlng to r vote dllutlon elalm ls twofold: to aseerteln whether mlnorlty
group membera eonstltute i polltleally eoheslve unlt and to determlne
whether whltes vo[e eufllclently er a bloe ueually to defeat the mltrorlty'r
preferred eandldate. Tlur, the quertlon whether a glven dlstrlet ir-
perlenees legally algnlfieent raelal bloc votlng requlrea dlserete lnqulrlea
lnto mlnorlty and whlte votlng practleea. A ehowlng that e algnifleant
number of mlnorlty group members urually vote for the aame eandtdates
ls one wey of provlng the polttleal eoheslvenese neoessarlf to e vot€i dltu-
tlon elalm, and eoneeguently eetablleher mlnoilty btm votlng wlthln the
tneanlng of ! 2. And, ln generat, i whlte bloe vote thet normally wlll
defeat the eomblned etrength of mlnorlty rupport pluo whlte.tto$over,,
votes rloea to the level of legalty elgnlf,eant whlte bloe votlng. Beeaure
loco of polltlcet power througfi vote dllutlon ls dlrtlnet ftom the merc tn-
ablllty to wln a pertleuler eteetlon, r pettern of raelel blm voUng tlret
extendr over a perlod of tlme ls morc probetlve of r elalm that a dlrtrlet
erperlenees algnlficant polerlzetlon then are the regult! ofa alngle elee-
tlon. In e dlstrlet where eleetlonl rre ehown uaually to be polerlzed, the
feet that raeldly polarlzed votlng h not present, ln one electlon or e few
eleetlons doer not neeesrdrlly negate the coneluslon that the dlctrlet er-
perleneee legally olgnlflcant bloc votlng. Furthermore, the oueeeas ofa
mlnorlty eendldate ln a partleutar eleetlon doea not neeeesarlly prove
that the dlstrlct dld not erperlenee polarlzed totlng ln that eteetlon.
llere, the lXatrlet Court'a epproaeh, whleh tested deta derlved lmm
three eteetlotr yeero ln eteh dl8trlct ln (ueetlon, end *hlch reveated that
blacka attongly iupported bhch eandldater, whlle, [o the blaek eendl-

THORNBURG u. GTNGLES llt

Syllabus

dates'ueual detrlment, whites rarety did, satisfaetorily arlrlresses eaeh
faeet of the proper standartl for legally aigriftcant racial bloc voting.
Pp. 19-27.

- 
2. The lang,age of $ 2 and its regisrative hretory prarnry demonstrate

that proof that some minority ."ndi.lat"e have been eleeted does not
foreelose a $ 2 elaim. T'lrus, the Distrlet court tlid not err, as a matter of
law, ln refuslng to treat the faet that some black eandi<tates have sue-
eeeded ae dispositlve of appellees' $ 2 elalms. where nrultimember dis-
trletlng generalty worke to dilute the minorlty vote, it eannot be de-
fended on the ground that lt sporadieaily and serendipitousry beneflts
tnlnority voters. Pp. dt-ne.

. 8. The elearly-erroneoue test of F ederal Rure of civir procedure 62(a)
ls the approprlate etandard for appelrate review of ultimate findings oi
vote dilution. As both amended $2 and its leglslative history riake
elear, in- evaluating a etatutory etaim of vote dilution through di.;ictin;,
the trial eourt ls to eoneider the "totatity of the eireumstinees', and io
d-etermlne, based upon a praeticat evarualion of the past and present re-
illtles, whether the politieal process is equally open to minority votere.
In thls ease, the Distrlct court earefully eonsidered the totaliiy of the
elrcumstanees and found that in eaeh distriet raeiaily polarized-voting;
the legaey of olflelal diaerlminatlon ln votrng matters, edueation, houi-
lng, emplo5rment, and health servieee; and the persistence o[ eampaign
appeals to raeial prejudiee aeted in coneert with the multimembei dis-
trleting. seheme to impair the ability of geogr"aphieally insular and politi-
eally eohesive groups of blaek voters to particifab equally in the political
proeess and to eleet eandidates of their cholce. pp. 4n_al.

Jusrree BnTNNAN, jolned by Jusrrcr Mensnill, Jusrrcr Bucx_
MUH, and Jusrrcs Srrvnus, eoneluded ln part III_C that for purposes
of 0 2, the legal eoneept of racially polarized votlng, as it retates io ciaims
of vote dllutlon-that ls, when lt ls used to prove [h"t the minority group
le polltlcally coheslve and that white votere wilr usually be able to alreai
the mlnorlty's preferred eandrdates-refers only to ihe exietence of a
eorrelation between the race of voters and the seiection of eertain candi-
dates. PIalntiffs need not prove causation or intent in order to prove a
prlma faele eaee of raelal bloe voting, end defendants may not retrut that
ease wlth evldenee of eausatlon or intent. pp. 2g-40.

. Jusnce BRENNAN, Joined by Jusncu Wrrrrr, eoncluded in part
IV-B, thet the Distrlet Court erred, as a matter of law, lrr ignoring the
eignlfleanee of the eustained suceess blaek voters have exp"erienci l in
Houee Dlstrlet,23. The perslstent proporfional representation for black
ieeldentE ln that dlstrlet ln the last six electrons iJineonsistent with ap-
pellees'allegatlon that blaek voters'ability ln thet distriet to eteet repre-
eentatives oftheir eholee is not equal to that e4Joyetl by the white major-



tv TH0hNBURG TI GINGLES

Syltabue

Ity. Pp. 43-4d.
iusricr O'Coxxoi, Jolned by Tnr CrrbrJusnen, Jusncr PowbLL,

and Jusncr RunNeutst, eoneluded that:
l. Ineofar is atetletlesl evldenee of dlveigeht raclat totlng petterhs le

ddrhltted eolely to eshbllah thet the mlnorlty group ls polltlcally eohe-

slve ahd to asseos lts proopech lor eteetorel BueGeBs, meh e rhowlrlg eah-

trot be rebutted by evldence that the dlvergent votlng patterns iniy be

bxplalned by eauses other then raee. Howevef, evldenee ol the reeaona

for dlvergent votlng patteini csh ln soine,elreumatanees be relevdnt td
the overatl vobe dllutlon lnqulry, end there ls ho rute egalnat eonglder'
itlon of etl evldenee eoneernlng votlng pteferenees other than rtstlstleal
evldenee ofraelal votlng patterna. Pp. l?-18.

. 2. Conslstent end auatelned aucceae by eandlda0es prefeircd by inlnof-
Ity votera la prerumptlvely lneonalatent wlth the exlrtenee of a ! 2 vlole-
tlprr. Ilre Dlstrlct Court erred ln aseeealng the bxtent of blaek eleetoral
suceecs ln House Dlstrlet $9 end Senate Dletrlet 22, as well es ln Houee

Dlotrlct 29. Exeept ln House Dlatrlet Zt, desplte these errors the Dls-
trlet Court'a ultlmete eonetuslon of vote dilutlon lo not ctearly el?oneoull.

But ln House Dletrlct 28 appellees falled to estebllsh e vlotatlori of 0 2.

Pp. 18-22.

BRENNAN, J., anhounced the Judgment of the Court and detlvered the
oplnlon of the Court wlth respeet to Perts I, II,-l,ll-A, III--B^, IV-A, and

V, ln whleh WnmE, ItlARsHAr,L, BlActrltur, end Srpvults, JJ., Jolned, an

oplnlon with respeet to Pert III-C, ln whlch MAnsnALL, Bt.AcKMtlN, and

Srevrxs, JJ., Jolned, and an oplnlon wlth respeet to Pert IV-8, ln whleh

WHItc, J., Jolned. YIHrfE, J., f,led a eoneumlng oplnlon. O'CoNNon, J.,
flled an opinlon concurrlng ln the Judgment, ln whleh Bunorn, C. J.,
PowEr,r., end Rnnweutsf,, JJ., Jolned. SrnvnNs, J., ffled an oplnlon

coneurrlng ln part and dlssentlng ln part, ln whlch MtnsrHlu and

BLAcKnuN, JJ., Jolned.

NOTICE: I1ls oplnlon h subject to formal revlaion before publlcatlon ln the
or:llmlnerv orlnl of the United Steter Reoortr. Readerc rre requealcd to
irotlfv the htiporter of Declelonr. Supreme Court ofthe United Stetls, Wash-
lnstdn. D. Cl mffS, of rn.y tyiogr'aphlcal or other formal errors, in order
thit conectlons mey be made before the prellmlnary print gms to presE.

SUPREil{E COURT OF TIIE T]NITED STATES

No. 83-1968

LACY TI. THONNBURG, ET AL., APPELLANTS U.

RALPH GINGLES ET AL.

ON APPEAL FhoM THE UNITED STATES DISTRICT COURT FOR
THE EASTERN DISTRICT OF NORTH CAROLINA

lJune 30, 19861

JusttcE BnsNHeH annolrneed the judgment of the Court
and delivered the opinion of the Court with respect to Parts
I, iI, III-A, III-8, IV-A, and V, and an opinion with respect
to Part Itl-C, in which Jusrrcn MeRsnall, Justtcn
BucrnuN, and JusrIcE SrEvENs join, and an opinion with
respect to Part IV-B, in which Justtcn WnItn joins.

This ease fequires that we eonstrue for the first time $ 2 of
the Voting Rights Act of 1966, as amended June 29, 1982.

42 U. S. C. $ l9?3. The speeifie question to be decided is
whether the three-judge Distriet Court, eonvened in the
Eastern Distriet of North Carolina pursuant to 28 U. S. C.

i2254(a) and 42 U. S. C. $ lg73(e), correetly held that the
use ln a legislative redistricting plan of multimember districts
in flve North Carolina legislative distriets violated $ 2 by
impairing the opportunity of black voters "to participate in
the politieal process and to eleet representatives of their
ehoiee." $2(b), 96 Stat. 134.

I
BACKGROUND

In April 1g82, the North Carolina General Assembly en-
acted a legislative redistricting plan for the State's Senate
and House of Representatives. Appellees, blaek eitizens
of North Carolina who are registered to vote, challenged



THORNBURG a GINGLES

seven disttiets, one slhgle-membert and six tnultiniembert
distriets, alleging that the redlstrieting ochetne impaired

blaek eiti,ens, ability to eleet fepresetr[atives of thelr ehoiee

ln vlolatloh of the Forrteenth and Flfteenth Amendments to
the United Shtes Constltution and of $2 of the 'Vofing
klghts Aet.t

Aft"r appellees brought suit, but befoie trlal, Congrese

amended 0i. Tte emendment was targely e negponse to thls

Court's plurality bplnlorr ln ltobil.e v. Boldm,4d6 U. S' 66

(1980), which had deelared that, ln ordet bo estirblish e vlola-

tion oi either $ 2 or of the Fourteehth or Fifteenth Amend-

thents, tnlnoiity votere trtust prove that e coritested eleetoral

meehanism *".lnt.tttionally adopted ot tnalntalned by stete

ofileiats for a dlseriminatory putpose. Coh$tess substeil-

tlaliy revlsed $ 2 to tnake elear that d violatlon eould be

proi"n by showing dlserimlnatory bfreet alone and 
!,o 

estSb;

ii.h 
"t 

the retevant legal standard the "result8 test," applled

by this Court inWhite v. Rqeslet, ll2ll. S. ?66 (1913), end

by other federal eourts before Bolden, suqru" S' Rep'

No. 9l-41?, P. I (1982) (herelnafter S. Rep')'

Seetion 2, as amended, 96 Stat. 134, reads as follows:

'Appettees ehallenged Senete Dlotrlct l.to. 2, whleh eoirslstcd oi the

*h;il; Northampto-n, Hertfotd, Gatec, Bertle, and Chowan Countlel'

";J;""t" 
of S'ashington, Mertltr, Hellfar, and- Edgecombe C9lntte1'_--'-dpptt"". 

challeigd the foltowlng multlmember diatrletr: Senate

No. iliUg.tlenburg-and Cabamrs CountlesJ membera), House No' 80

(Mecklenburg County--€ members), House No' 39 (part of Forayth

E;ilyi i".b"*i, House No.23 (Durham Countv-J P"IP"")'
iorr" No. 2l (Weke County-4. members), and Houee No' 8 (mlgon'

Naah and Edgeeombe CountleoJ membere)'- ;epp-ff*6 lnttlaterl thle eetlon ln September 1981, ehallengllg the

Nortir'Cerollne General Aasembly,s July l98l rerlistrletlng. The hlstory

ofthlgaetlonlsrecountedlngreeterdetalllntheDlstrletCourt'goplnlonln
[ftt .""", Gin4lec t. Dilmbin,690 F. Supp' 346, 350'-368{EPNp 1984-)'

il;r;n.!,;o[; note that theGeneral Assembly revleed the lg6l pletr ln

l-p;ii 1982 and that the ptan at lssue ln thia ease ls the 1982 pleh'

THORNBURG U. GINGLES

"(a) No voting qualifieation or prerequisite to voting or
standard, praetiee, or proeedure shall be imposed or
applied by any State or political subdivision in a manner

whleh results in a denial or abridgement of the right of
any cltizen of the United States to vote on aeeount of
reee or eolor, or in eontravention of the guarantees set
forth in seetion 4(fX2), as provided in subseetion (b).

"G) A violatioh of subseetion (a) is established if, based

oh the totality of the cireumstanees, it is shown that the
politieal proeesses leading to nomination or election in
ihe State of politieal subdivision are not equally open to
partieipation by members of a elass of'eitizens protected
by subsectlon (a) in that lts members have less opportu-
trity than other members of the electorate to participate
in ihe politieal proeess and to eleet representatives of
their ehoiee. The extent to which members of a pro-

teeted elass have been elected to offiee in the State or
politieal subdivision is one cireumstanee which may be

eonsidered: Prouided, That nothing in this seetion

establishes a right to have members of a proteeted elass

eleeted in numbers equal to their proportion in the
population." Codified at 42 U. S. C. $ 1973.

The Senate Judieiary Committee majority Report ac-

colnpanlng the bill that amended $ 2, elaborates on the

elrertmstances that inight be probative of a $ 2 violation, not-

lng the following "typieal faetors:"'

"1. the extent of any history of offieial discrimination in
the state or politieal subdivision that touehed the right of
the members of the minority gfoup to register, to vote,

or othei'wise to partieipate in the democratic process;

,These faetore were derived from the analytleal framewot'k of wh.ile v.

Regester, 412 lJ. S. ?56 (19?3), as reflnetl antl developetl try the lower

eourts, ln partietrlar by the Fifth Circuit in Zirttmer v. lle'Ikithert', 486

F.2d l*gi (1913), atT'd gub nom,. East Carmll Parish' School Board v'

Iiarshnll,424 U. S. 636 (19?6) (per curiam). S. Rep' 28' n' ll3'



THORNBT,RG O. OINGLES

"2. the extent to whieh voting in the electlons of the

state or politieal suMivlslon is raeially polatized;

"3. the extent to which the state of political suMivlsion
has used uhusually large eleetion districts, tnqJority

Vote tequirementsi entl-iingle shot ptovisions, or othei

ilidit".ti."s oi proeedires thal.may gnhatr-cu lf"
opprfrrnty for diserimination against the minorlty
group;

"4. if there is a eandldate slating proeess, whether the

members of the mlnority group have been denied aeeess

to that Proeess;

"6. the extent to whieh members of the minority group

in the state or politiea! subdivision bear the effects of

diserimination in such areas as education, employment

"nJ 
fre"fth, ivhleh hinder their ability to partielpate

effeetively in the politieal proeess;
,,6. whether politieal earnpalgns have been eharaeterized

by overt or subtle raelal aPPeals;

"?. the extent to whieh members of the minority group

have been elected to publie offiee ln the jurlsdietion'

."Additional faetors that in some eases have had prola-

tive value as psrt of plaintiffs, evidenee to establish e

violation are:
,,whether there ls a eignifieant lack of responsivenesa on

the part ofeleeted ofnelals to the particularlzed heeds of

, the members of the hinority group'
,,whether the poliey underlying the state or politieal strb-

airisioJs use of eueh voting qualifteation, prerequlsite to

;rfittg; ;" tt nd"td, practiee or proeedure is tenuous'"

S. Rep. 28-29.

TheDistrletCourtappliedthe..totalityofthecircum-
rt"n."t;l"tt."t forth in iz(t) to appellees'statutory elaim'

lrJ, t"fvirg prineipally on the facbors outlined in lh:Ilnai:
ffiori;';Ai hJG. redisttietlng :.h.:f" yl*1-13
;;;;; i[ i"trrlJ in ihe ditution of blaek citizens'votes in all

THORNBURG u. GTNGLES 6

seven disputed distriets. In light of this eonelusiotl, the
eourt did not reaeh appellees'eonstitutioiral claims. Gingles

v. Edmisten, 690 F. Supp. 346 (EDNC 1984).

Preliminarily, the eourt found that black eitizens consti-
tuted e distihet iropulation and registered-voter minority in
eaeh ehallenged distriet. The court noted that at the time

the multimember districts were ereated, there were con-

eentrations of blaek citizens within the boundaries of each

that were sufffeiently large and eontiguous to eonstitute
effeetive voting majorities in single-member distriets lying
wholly within the boundaries of the multimember districts.
With respeet to the challenged single-member district, Sen-

ate Distriet No. 2, the eourt also foutrd that there existed a

eoneentration of blaek citizens within its boundaries and

within those of adjoining Senate District No. 6 that was suffi-

eient in numbers and in contiguity to eonstitute an effective
voting majority in a single-member tlistriet. The District
Court then proceeded to find that the following eircum-

stahees eombined with the multimember districting scheme

to result in the dilution of blaek eitizens' votes.
First, the court found that North Carolina had officially

diseriminated against its blaek eitizens with respeet to their
exercise of the voting franchise from approximately 1900 to
l9?0 by employing at different times a poll tax, a literacy
test, a proliibition against bullet (single-shot) voting' and

'Bullet (slngle-shot) votlng has been described as follows:
,,.Conslder fel town of Gfi) whites and 4fi] blacks wlth an at-large election

to choose four eouncit members. Eaeh voter is able to cast four votes.

suppose there are elght white eandldates, wlth the votes of the whites split

among them approxlmately equatly, and one black eandidate, with all the

blacks'votlng for trt, and no one etse. The reeult is that each white eancli-

da[e reeelvee about 3fi) votes and the black eandidate receives 4(x) votes.

The bteek hee probabty won a seat. Thls teehnlque ls ealled single-shot

votlng. Single-shot voting enables a minority group to wln some at-larg.e-

&eblf lt eoneentrates lts vote behind e llmlted number of candidates and lf
the tote of the rn{orlty is dlvlded smong a number of eandidates" " City
of home v. lJnitctl Stalea, 446 U. S. 166, 184, n. 19 (1980), quotlnC U' S'



THORNBURG U. GTNGLES

designated seat ptsns' for multimetnber districts. The

eourt observed that even after the removal of direet barriers
to blaek voter registratlon, eueh as the poll tax and literaey

test, blaek voter registration remained relatively depressed;

in lbgz only 62.'l% of age-qualifted blaeks statewlde were

registered to vote, whereas 66,7% of whites were registepd'.
Th"e Distriet court found these statewlde depressed levele of
blaek voter registratlon to be present in all of the diaputed

distriets and to-be traeeable, aLleast ln part, to the historleal

pattern of statewlde offleial diserimination'
second., the court found that historle diserlmination in

edueatiorl housing, employment, and health serviees had

i".rGa h a lowei'soeloeconomle status for North Carolina

bia.ks as a group than for whites. The eourt eotreluded tha[
thls lower rt"tur both gives rise to special gfoup lnterests

nna ntna"rr blaeks, ability to partielpate effeetlvely in the

potitt""t process and to eleet representatlves of thelr ehoiee.

Thid,, the eourt eonsldered ither voting proeedures tha!'

mdv oDerate to lessen the opportutrlty of black voteis to eleet

""naia*"r 
of thelr cholee. 

- 
it noted that North Catolina has

;;ffiiat vote tequlrement for priqarfl eleetlons dhd, while

a.iiJi*i"i',rg lhai rio btaek candidate for eleetlon to the

d;l; C"iti*fAtr".bly had failed t'o win solelv beeause of

thls tequlrement; the corlrt coneluded that lt nonetheless

bresetrts a contlnulng practleal ltnpediment to.the oppo!'q-

IiW "i 
fU.k vothglninorlties to eleet eandidates of thelr

eholee. The eoUrt 
"iro 

r"."rked on the fact thet t'lorth

drioiin" does not have I subdlstrlct resldeney requlrement

i* membert of the General Assembly elected from

multim"mber distriets, a reqtrlrement which the eourt fouhd

Commlaston on Clvlt Rlghh, fire Votlng Rlghts Aeh Ten Yeare Aft'er'

pp. 206-2{n (19?6).
"jn"rtgn"t"a (or numbercd) eeet echentes rcqulre a eendldate lor elee-

tl;il;;ttl;emnet art-rrJr 
" ryl tor.tpetiRe reati, and ednt*rtdet

;rtrt" elreutrlstaricel, f*ttrat" bullet votlng" Se6, a' g'' Clll oI Edonl'

tnpn| * 186' ii. 21.

lr

TIIORNBURG o. GINGLES 7

eould offset to some extettt the disadvantages minority voters
often experienee in multimetnber districts.

Fm,rth, the eourt found that white eandidates in North
Carolina have eneouraged voting along color lines by appeal-
ing to raeial prejudiee. It noted that the record is replete
with speeiffc examirles of racial appeals, ranging in style from
overt and blatant to subtle and furtive, and in date from the
1890's to the 1984 eampaign for a seat in the United States
Senate. The court determined that the use ofracial appeals

in politieal campaigns in North Carolina persists to the
present day and that its eurrent effeet is to lessen to some

degree the opportunity of blaek eitizens to participate effec-
tively in the political proeesses and to elect candidates of
their choice.

Filth, the eourt examined the extent to which blacks have

been elected to offlce in North Carolina, both statewide and

ln the ehallenged districts. It found, among other things,
that prior to World War II, only one black had been elected
to publie offlee in this century. While recognizing that "it
has now beeome possible for blaek eitizens to be elected to of-
fiee at all tevels of state governmeht in North Carolina," 590

F. Supp., at 367, the eourt found that, in eomparison to white
candidates running for the same offiee, blaek eandidates are

at a disadvantage in terms of relative probability of sueeess.

!t also found that the overall rate of black eleetoral stteeess

has been minlmal in relation to the percentage of blacks in the
total state population. For example, the court noted, from
l9?1 and 1982 there were at any given time only two-to-four
blaeks in the l20-member House of Representatives-that
ls, only l.6Vo to 3.37o of House members were black. From
l9?5 to 1983 there were at any one time only one or two
blacks in the 60-member State Senate-that is, only ZVo to 4Vo

of State Sehators were blaek. By contrast, at the time of the

District Court's opinion, blacks eonstituted about 22.47o of
the total state population.



THORNBURG u. GINGLES
TIIOBNBURG I,. GINGLES

With respeet to the Buecess in this eentury of blaek ealdi-

dates iri the eontested distriets, see also Appendix B, infra,

it 
" "rurt 

found that only one blaek had been eleeted to House

Dirtri.t 36-after thls lawsult began. similarly, ohly one

[U.L .f,"a eerved ln the Senate from Distriet 22' from

19?6-1980. Before the 1982 election, a blaek was eleeted

;i; ft;; to the House froni Disttict 3e (part of Forsvth

6ri"ivf f, the 1982 eontest two blaeks were elected' Sinee

fyrd 
" 

Uin.t eitizen had been eleeted eaeh 2-year t"-ry t'lh:
House from Distrlet 28 (Durham County), but no blael( had

i""" 
"f".t.a 

to the Senate frorn DUrhain County. In House

birt"i.t 2l (Wake County), a blaek had been eleeted twice to

tf," Hoor", and another f,i"tk t"t'"d two !'ery1in !!re 
State

6;il;- iio blaek had ever been eleeted to the Hottse or

S"lrri.ft* the arei eovered by House Distriet No' 8 and ho

ffi;;;;;;;J;;;b""n elected to the senate from the

ar"".orut"d by Senate District No' 2'*Th;;;aia 
a"r.no",ledge the improved sueeeE' of blaek

."ndidrt", irr the lggZ eleittond, ln whieh 11 bleele were

;i;"hd to the State tloor" of Representatives, inelrtding 6

tir.t, froni the multimembet dtstiiets at issue here. How-

"r"i,-fft 
eourt pointed out that the 1982 eleetion wa8 eon-

dueted after the eommeneement of thls litigation' Th9-eou5t

r*natheeircumstaneesofthelg82eleetlonsufficientlyab-
erraHonal and the;;; by blaek candidates too minlmal

"iJ 
tr" reeent ln relation to ttre long history of eomplete

denial of eleetive opportunitles to eupport the eoneluslon that

blaek voters, opporiunities td eleet representatives of their

ehoiee were not imPairgd' . i. : .. ..

Fianltg,tt 
" 

eouii eonttd"t"a the exterrt to whieh voting in

th;;;li'Ajed districh lryas reeiallv poiarlzed' Based on

,t tirii."t &id"n " 
presented by expert witnesses, aupple-

mehted to some degr'eeby the b;tim;ny of lay wltnesses' the

court found that a[of the ehallenged distriets exhlbit severe

and persistent raelally polarlzed votlng'

Based on these findings, the eourt rleelared the contested

portions of the 1982 redistricting plan violative of $ 2 and en-
joined appellants from condueting elections pursuant to those

portionJ of the plan. Appellants, the Attorney General of
iVorth Carolina and others, took a direet appeal to this Court,
pursuant to 28 U. S. C. $ 1263, with respeet to five o! the
multimember dlstriets-House Distriets 21, 23,36, and 39,

and Senate bistrict ZZ. Appellants al'gue, first, that the

Disttlet Cottrt utilized a legally ineorreet standard in deter-
mining whether the eontested districts exhibit raeial bloc vot-

ing tJan extent that is cognizable under $2. Second, they

eontend that the eourt used an ineorreet definition of racially
polarized voting and thus erroneously relied on statistical
evidenee that was not probative of polarized voting. Third,
they maintain that the eourt assigned the wrong weight

to evidence of some blaek candidates' electoral sueeess' Fi-
nally, they argue that the trial court erred in concluding that
these tnultlmember distriets result in black citizens having

less opportunity than their white counterparts to participate

in the politieal proeess and to eleet representatives of their
cholee. lVe troted probable jurisdiction, 471 U. S' 

-(1985), and now affirm with respect to all of the districts
bxeept House District 23. With regard to District 23, the
judgment of the District Court is reversed and remanded for

further proeeedings 
II

SECTION 2 AND VOTE DILUTION THROUGH USE
OF MULTTMEMBER DTSTRICTS

An understanding both of $2 and of the way ih .whieh

multimember distriets ean operate to impair blacks' ability to

eleet representatives of their ehoiee is. prerequisite to an

evaluation of appellants' eontentions. First, then, we re-

view atnended $ 2 and its legislative history in some detail.

Second, we explain the theoretieal basis for appellees'claim

of vote diltttion.



THORNBURG U. GINGLES ll
Io THORNBUhG U' GINGLES

A

SECTION 2 AND ITS LEGISLATIVE HISTORY

subsectiotr 2(a) prohibtts all states and politieal subdi-

,irioft. frorn lmposin1 anil voting qtialiflcatiohb or prereq-

;i;il; b votlng, or a-ny sfendardt, practicesl ot.p:..9t:dl1fl

whteh result ln the denial or abrldgment of.the rlght. to vote

oi nny eltlzen $ho ls a metnber of a protected class ot raetal

;il1i"*G; ,inoritl"r. subsectio, ztt-l establishes thar $ 2

;; il; vlohted where the ,,totality of the eireumstahees"

;;;[h"t;.tt e potiti.al ptocesses.leading to nominatign or

eleetlotr . ., . 8f€ not equally opert.to participatlo'PL TiI:
bers of a [proteeted el*ssl . ' ' lrt that its membets have le88

ffi*uii,Irih"n otfrJi in"rrbers of the eleetorate to partiel-

oare ln rhe potitieai;;;;a to "i"* 
r"presentatlvee.of

lheh choiee.;' While explainlng that "[tlhe extgnj t'o whlcn

ttr"tnU"t. of a proteeted class heve been eleeted to olllee ln

il;'$1ri. o" pontifi-su6ivlslon is ond eireumstanee whieh

il;y b;;;ii"r;;i, evaluating an- alleged vio-latlon' $.2(b)

.ffiirrr-},at ,,nothing in t$21 establlshes a rlght to have

."ri"t. of a protect-ed elass- eleeted in numbert equal to

their proportioh ln the population'"-'tirJ SJt"te RepoJ if,i.h aeeoinpahled the 1982.amend-

tnehts elaboiutes 
"n 

if,e n"ioie of i Z vtolatlohs and on the

iioo[r"ouired to 
"rt"ttirtt 

these violailons.' First and fore-

ffi";#'iilpq'fiiilosltlvelv l'91':!' the positiol'I.th''

tir.irW'if-iiiiiir- i-' Botdnn, 446 u' s' 66 (1e80)' ivhlch

required proofthat the eontested electoral practice or.meeha-

niJn, *"" adopted or maintained with the intent to discrimi-

nate against minority voters.s See, e' g', S' Rep' 2,15-16,
27. T-he intent test was reputliated for three prineipal rea-

sons-it is,,unneeessarily divisive beeause it involves charges

of reeism oh the part of individual offfcials or entire eommuni-

tles," it plaees an "inordinately diffieult" burden of proof on

pfJirtiffi and it "asks the wrong questlon"' Id', at 36' The
ltigttt" question, as the Report emphasizes repeatedly, is

whitheri,as a result of the challenged praetice or structure

ftrintifft do trot have an equal opp-ortunity to partieipate in

[t " 
prtiti."l pioeesses and to eleet candidates of their

choiel."' kt., aL28. See also id., at2,2'1,29, n' 118, 36'- 
in otd.t to answer this question, a eourt must assess the

lmpaet of the eontested strueture or practice on minority

.i"ltor"t opportunities "on the basis of objective_ {a9tors.."
ii., 

"t27."T1're 
Senate Report speciftes factors whieh typi-

ealiy tnay be relevant to a $2 claim: the history.of voting-

terat"a d-iserimination in the state or political subdivision; the

"i["nt 
to which voting in the eleetions of the State or political

, The Solleltor Cenerel urgee thls Court [o give little welgh[ to th-e- Sen-

*"'ri"plJ, 
'"igrlig 

ttt"i f r"presents d eompromlse among confllet'lng

"i*ti'rir,; ffi-t'uJls **"ho-ri 1".. authorlhllve than mort commlttee

d;,tu.' siLi lrt unttJiLt"t as Amiana Curioe 8' n-' l.2l ?'": ls'
i'#;;r, *;;J ;;* 

" 
t"gt.r"Ur" htetory of emended 0 2 eontalns

anything to lead us to **i'a" tftit thls Senate li-engrt ahould be aeeorded

iiir"'*?rlit. w" rtl," rep"itedlv reeognlzed thet the authorltatlve

eource for leglalatlve t,;;d;il th;-tonin,ltt"" reoorts on the bill' see'

c. o. Gorcia v. llnibl Slnhc,469 U' s' ilNlSa]r zubn v' Allen'

mo u. s. 168, t86 (1969)'

!T'he Senate heport states that amended !2 was designed to restore

[hei,resultg test"-the legal standard that governed voting discriminatign

c""", pilo. to our deelsion ln Mobila v. Boldan, 446 U. S. 66 (1980).

i. n"p. ff-fe. T'he Report notes .that ln.pre-Bold'en easeg sueh as

irgiirr,4lz U. S. ?66 (10?3), and Zimmnt,486-F' 2d 129? (19?3)' plain-

6d eould prevall by ehowing that, under,the totelity of the elrcumstanees,

"..f,"tilil;leetlon 
taw or frocedure had the elfeet of denying a proteeted

minortty-"n equal ehanee to partlclpate ln the-electoral process. under

li;;;rlt test," plalntiffs-are not required to demonstrate that the

"il"1"rg"6 
eleetorat'law or strueture waa deslgned or malntained for a

dlserlminetory purpose. S. Rep. 16.- ift" SeneL'Committee lound ihat "voting praetices and proeedures

that, have dlserlmlnatory results perpetuale the effecta of past purposeful

dlserimination." S. Rep. 40 (footnote omitted)' As the Senate Report

hotes, the purpose of the Voting Rlghts Act w-a8 "'not only to eorreet an

J"tl"" t trto.y of dlacrimlnatlon, the denying to Negroes of the right to reg.-

ilt"r 
"nO 

vo[e, but also to deal with the aeeumulatlon of diserimlnation."'

S.R"p.6(quotlnglllCong.Ree.S2g6(1965)(remarkeofSen.Javlta)).



subdivision is raeially polarized; the extent to whieh the state

or politieal subdivision has used voting praetices or 
-proee-

dures that tend to enhanee the opportunity for diseriminatlon

nn"i"tt the minorit, ErouP, sueh as unusually large electlon

dlstriets, majority vote requlrements, and pronlblt'lons

"g"..i 
irftet votins; the exeluslon of members of the minor-

it] gtoup from 
"aididate 

slating proeesses; the 
-extent-to

*if&t minority group members hear the bffeets of past dls-

erimination in areas sueh as edueation, employment,.and

[""'ftft, *f,i.ft hindet their ability to pattieipate ef3ciivetl-i,1

the poiitieal proeeEs; the use-of overt or subtle raeial appeals

i;6li;""1 eampaigns; and the extent to whieh members of

the minority Srbup have been eleeted to publie office inthe

ir.rrJi.i.r. i. ri"p. u-?!;see also tury: "t.=^ :, Tl
Report notes also that evidenee demonstrating that ele-eted

offleials are unresponsive to the partieularlzetl needs of the

*.*f"* of the minority group and that the poliey- underly-

in* tt " 
State,s or the potiti."i subdivision's use of the eon-

teited praetiee or structure is tenuous mfy have Probative

,"tu". S. Rep.29. The Reportstresses, however' thatthls

ii.iri typical iaetore ls neither eomprehensive rror excluslve.

While thl 
"nu."rated 

factors rt'lll ofteh be pertinent-!9 ::t-
il;'iil, ,i sz violations, partieularly to vote dllu6on

.i"ir.,; other factors may also^be retevant and may be eon-

;ilffi. 
-s. 

n"p.29_g0. Furthermoie, the senate commit-

L" onr"*"a that "there lS ho requirement thdt ahy paftlctt-

i;;;;;1"" or rr.tott b" p-v"d, or that a rnqforitv.of theni

ilril "* *av or the other.' Iil', at29' Rather' the com-

I"d""'l;;;"r""J' ir,"[ ;rhe quesrion wherher the political
-[io.g."", 

are 'equally open' d"i"ntls upoh 0 searthitrg praetl-

eal evaluatiort of d;';;;t eni pt"t"nt feality"".id'1 at.30

(footnote omitterl),'Jnion 
""run.tlonal,, 

tie* of the polltlcal

pfoeesE. Id., eil80, ft. lm'

THORNBURG U. GINGLES

rsection 2 prohlbits all forms of votlng diserlmlnetlon, not Jrtst vote

rlilutton. S. ReP. 30.

TIIORNIItIRG u. CINGl,li)S

Although the Senate Ileport espotlses a flexible, faet-

intensive test for $ 2 violations, it limits the eireumstanees

under whieh $ 2 violations may be proved in three ways'

First, electoral devices, sueh as at-large eleetions, may not

be considered per se violative of $ 2. Plaintiffs must demon-

strate that, under the totality of the cireumstanees, the

deviees result in unequal aeeeEs to the eleetoral proeess.

1d.., at 16. Seeond, the cor{unction of an allegedly dilutive

electoral mechanism and the laek of proportional representa-

tion alone tloes not establish a violation. lbid. Third, the

results test does not assume the existenee of raeial bloe vot-

ing; plaintiffs must prove it. Id., at 33.

B

VOTE DILUTION THROUGH THE USE OF

MULTIMEMBER DISTRICTS

Appellees contend that the legislative tleeision to employ

muliimember, rather than single-member, districts in the

contested jurisdietions dilutes their votes by submerging

theminawhitemajority,r'thusimpairingtheirabilityto
eleet representatives of bheir ehoiee."

l3

" Dilutlon of raeial minority group voting etrength may be eaused by the

dlspersal of blacks lnto dietiicts ln whlch they eonstitute an ineffective

.iro.ity of votere or lrom the concentration of blacks into districts where

thev constitute an exceeslve maJorlty' Engstrom & Wildgen, Pruning

it o.n. fro'n the fiieket: An Empirlcal Test of the Existence of Racial Ger-

hymanderlng, 2 Legis. Stud. Q. 466, 466-466 (Iff?) (hereinafter Engstrom

d WifJg""l]' Seelko Derfner, Raciat Diserimination and the Right to

Vrt", ZE Vand. L. Rev. 623, 653 (19?S) (hereinafter,Derfner);..F' I"tl:t'
n".i"f C"""Vmanderlng and Legislative Reapportionment (hereinafter

parL"r), ln llilnority Vote Dilution86-100 (Davklson ed,, 1984) (hereinafter

tlinorlty Vote Dllution).

"The elalm we addrese in this oplnlon is one ln which the plaintiffs al-

UgJ 
"na 

attemptetl to prove that their ebility to electLhe representatives

of-thelr ehoiee was impaired by the seteetion of a nruttimember electoral

strueture. we have no o..r.ion to consider whether $ 2 permits, and if
t ao"t, what standards shoul(l pertain to, a claim brought by a minority

group,'whieh ls not sulllciently targe and eompact to eonstitute a majority



THORNBURG U. GINGLES

The essenee of a $ 2 elaim ls that a eertain eleetoral law,
praetice or structure lnteraets with social and historieal eon-

iitions to eause an inequality in the opportunities eqioyed by
blaek and white voters to eleet their prefemed represent-

atives. This Court has long teeognized that tnultimentber
distrietg and atJarge votlng schelnes mey "'operate to mini-

mize of eaneel out the voting strength of raelal [mlnorltles lnl
the Voting populatlorl."'rr Bu,rns v. hinh,ardao?r, 384 U. S'

ln i slngte-member dlstrlct, alteglng that [he use of a multlmetnber dlotrlet

lmpalrs lts ablllty ln itrfiuauc etettlono.
irye note also tlat we have no oceaslon to eonelder whether the stendards

we Bppty to ieapondenta' elaltn that multlmembet dlstrlets gperele to

aliu[e lfrl vo1" of geogrephleally ohealv6 mlnorlty group., whleh erelarge

enough 0o eonstltute mqlorltles ln rlngle-membet dlrtrleti atd whleh are

eontafned *lthln the bounderler of the ehsllenged multlmember dlatrlets,

are fulty pertlnent to other aortr of vote dltutiort ctalms, such es t clalm

alleflnj drat the apllttlng of a large end geographleally eoheslve mlnorlty

between two or more multlmember or elngle-member distrlets resulted ln

the dllutlon of the mlnorltY vote.
rCommentatofo sre lrr wldeepread agreement wlth thls eoneluslon.

See, c. g., Berry & Dy", Tlre Dtscrlmlnatory Ellecta of At-Large Elee-

itonr, Z-fi". St. U. L. frev. 8E (1919) (herelnafter Berry & Dye); Blaeksher

C U"n"t"", From Beytaldc v. Sinu la City of Mobile v' Bold'en, 8l
Hastings L. J. I (198L) Gerelnafter Blaeksher_ & Menefee); Bonapfel,

Ulnortly Challengec to At-tarSp Etectlons: Tte Dllutlon Problefn, l0 Gd.

L. n"" s6S (19?6i GerelnafteiBonapfet);Butler, Conatltutlonal and Stat-

utorv challenses to Eleetlon struetures: Dllutlon and the value of the

nts6t to VoteI4Z La. L. Rev. 861 (l9SZ) Oerelnefter Butler); Cerpenetl,

L"'gl.Uif r" Apportlonmenh ltlultlmember Dletrlcts and Falr liepreaente-

ffi;Dg pe.-L. Rev. 660 (19?Z) (heretnefter-Carpenetll.Davldron &

KorLl, At-Large Electlona and ltlnorlty Group Repreeehtltion, ln-Illlnor-

Itv Vote Dllutlon 65; Derfner; B. Grofman, Alternatlv* to slngle-llember
piurallty Dlstrlets: Legat and Emplrleal ksues (herelnafter Grofman,

At;;;L;tr,ln Reprientatlon and Redlstrictlng Issueo lO'il (Bt Glof;

m"n, n. fiiphart,. iI. M.K"y, & H. Searow eds', 1982) 
-(herelnafteh

n"pi"r"nt"tion ,nl R"dlrt.leilng laaues); Hartman, Raclal Vote Dllutlon

"ri-S"p"*tl"n 
of Powera, 60 G;: Waslr' La-nev' 080 (19S2); Jewell' The

Con."qr"r."" rif Slngle- end Hultlmember Dletrletlng, ln Representatlon

ind Reitstrl.tlng losuer iZg (tgSZ) (herelnafterJewell);Jones, !1e Imna3.

oil*"f Elee6on-Syatemr on Foltilcat Repreaentatlott, ll Urb. Afi. q. 3fr

THORNBURG U. GINGLES

?3,88 (1966) (quoting Fortsonv. Dorsey,379 U. S' 433' 439

(1966)). See also Rogers v. Lodge,458 U. S. 613, 617 (1982);

White v. Regester, 412lJ. S., at 765; Wltitcomb v. Chnttis,

403 U. S. 124, 143 (19?l). The theoretical basis for this type

of impairment is that where minority and majority voters
eonslitently prefer different eandidates, the majority, by
virtue of its nrlmerieal superiority, will regularly defeat the
ehoices of minority voters.lr See, e. g., Grofman, Altema-
tives, irt Representation and Redistricting Issues ll3-114.
il{ultimember districts and at-large election sehemes, how-

ever, are hot, per se violative of minority voters' rights.
S. Rep. 16. Cf. Rogers v. Lodge, supra, at 617; Regester,

sttytu, at ?65; firhitcomb, supra, Lt 142. Minority voters
whro eontend that the multitnember foim of districting vio-

lates $ 2, fnust prove that the use of a multimember electoral

strueture operates to minimize or eaneel out their ability to
elect their preferred eandidates. See, e' 9., S. Rep. 16'

While many or all of the factors listed in the Senate Report

may be relevant to a elaim of vote dilution through submer-

g"n." in multimember districts, unless there is a conjunet'ion

of tn" foilowing eireutnstances, the use of multimember dis-

triets generally will not impede the ability of minority voters

to eleet representatives of their ehoice.'6 Stated suecinetly,

(1976); Karnlg, Blaek Resourees and City Council Representation, 4l J'
Pol. lha (l$29); Karnls, Btack Represen[ation on City Councils, 12 Urb'
Afr. Q. 28(1s76)i Parker 8?-88.

,, Not only does "[v]oting along raclal lines" deprive minority voters of

their preferred fepresentative in these eircumstanees, it also "allows those

elected [o ignore [mlnority] lntereste without fear of political conse-

quenees," RoSnt v. I.adge,46S U. S. 613, 623 (1982), leaving the ririnority

effectlvely unrepresented. see, e. g., Grofman, Shoukl Representatives

Be Typlcal of Their constituents?, ln Representation and Redistricting

lssues'9? (hereinafter Grofman, Shoukl Representatives be Typieal?);

Parker 108.
[Under a "funetional" view of the politlcal proeess mantlatetl by $2'

S. Rep. 30, n. 120, the most important Senate Report factors beal'ing on

$ 2 challetrges to muttimember dietricts are the "extent to whieh minority

gtoup membere have been eleeted to public o(flce ln the jurisdiction" and



.,, i,,.ri.,;ii,:iiili l.*:j,rl$.ff*Hl fhtrglhu[fil}lii{aiEfi$re 1

tG THOBNBURGu GTNGLES

a bloc voting rrrqiority mubt t s?rorly be able to defeat eandi-

dates supported by a potitleally eohesive, geographlcally ln-

sular minority group. Bonapfel 366; Blaeksher & Menefee

34; Butler 903; Carpenetl6g6-699; Davidson, Minority Vote

Diiution: An Overvlew (hereinafter Davidson), in Minorlty
Vote Dilution 4;Grofman, Altefnatives 117. Cf. Bold,en,446

U. S., at 105, n. 3 (MeflsHALL, J., dissentlng) ("It is obvious

the,,extent to whleh voting ln the eleetlone ofthe stste or polltleal subdi-

vlslon ls raclalty polarlzedl' Id,, Al-n, lf present, the other faelora,

gueh as the llnierlng elfeeh of paat dlecilmlnatlon, the use of appeals to

raelel blas ln e-ieetlon eatnpalgnt, and the use of electoret devlees whleh

enhance.the dllutlve efects of tnuttlmember dlstrleh when aubetantlal

whlte btoc votlng dxlste-foi example entibullet voting tews and mqforlty

vote requlrementa, are eupportlve of, but not essenlial lo, a mlnotlty
voter'a elalm,

ln reeognlzlng the[ some senate Report faetora atre ,nore lmporta-nt to

rtulilmem-ber distrlet vote dllutlon elelme than others, the Court effectu-

ates [he lntcnt of congreas. lt ls obvlous that unlees mlnorlty group

merhbera expeflenee rotut"ntt"l dlffieulty electtng repreaehtatlvee of l,helr

eholee, they eennot prove that e challenged eleetoral meehenlsm lmpalre

if,"f. lmfW ,,to etit." !Z(b). Andf where the eontested eleetorel

st"uctu." l. a irluttlntember dlatrlet, eommentators and eourts agree that ln

ti* 
"Ur"n." 

of eignlflcerrt whlte bloc votlng lt eannot be aald thet, the eblllty

oi minorlty votei to eleet thelr ehoeen representatlves ls lnferlor to that of

whlte voter:s. See, e. g., McMiUanv. EacombiaCounlg, T4S F' 2d lGl?,

roai tcm ws4lt ilniiei St,,tet v. llanngo County Corntn'n, ?31 F' 2d

1646, 1566 (CAfl 1981) appeat dlsmlesed, eert. denled, 469 U' S' 
-(198i); Neustl v. Sidao, 6?i F. 2d 2og, ng (CAE l9?8), eert' denled, d46

U.-S."Sff (1980); Joitson v Halif,a'a County,69l F' Supp' 16l, l?0

innNC l9d4); tiiacksher & Menefee; Engetrom & Wldgen 469; Parker

ioz. Consequently, lf dilfieutty ln electlng and whlte btoe votlng are not

proven, rnlnoiity voters have not entabllshed thet the multlmember struc-

lure lnierfere, *ith th"l" abltity to eleet thelr preferred eandldates. Ml-

r,oiity rot"rr ray be able to proye that they stlll suffer aoeiel and eeonoinle

"ff*[ "f 
past diectltnlnatlon, thet appeals to raetal blas afe Employed ln

;i;;r" ffi;"rgnt, 
"nd 

that a rneiority vote ls requlred to wln e ry1t' lyt
they have trot d'emonstrated t substantlat lnabllity [o eleet eaused-by the

,r" or 
" 

multlmamber dlstrlet. By recognlzlng the prlmaey of the hlstory

and extent of mlnorlty eleetoral eueeeea tnd of raelel blm lotlng, the court
ilpi, i"qul.e. that-;Z platntlffe prove thelr ctalm before they rnay be

awarded rellef.

TIIORNBURG u. GlNGLIdS l7

that the greater the degtee to whieh the electoral minority is

homogeneous and insular and the greater the degree that
bloe voting oeeurs along majority-minority lines, the greater
will be the extent to which the minority's voting power is
diluted by multimember distrieting"). These eircumstanees

are neeessary preeonditions for multimember districts to
operate to impair minority voters'ability to elect represent-
aiives of their choiee for the following reasons. First, the
fninority group must be able to demonstrate that it is sufft-

eiently large and geographieally eompact to constitute a ma-
jority in a single-member distriet." If it is not, as would be

lh" ""*" 
in a substantially integrated district, Lhe mu,lti'

member form, of the district eannot be responsible for minor-
ity voters' inability to elect its candidates." Cf. Rogers,

t. ln thls ease respondents allege that within each eotrtestetl multi-
member dlstrlet there exists a mlnorlty group that ls sulficiently large and

eompaet to eonstitute a slngle-member district. In a different kind of

.r"", fot example a gerrymander case' plaintiffs might allege that the

mlnority group that ie sulllciently large and eompact to constitute a

alngle-member distrlet has been split between two or more multimember

or 
-elngte-member distriets, with the elleet of diluting the potential

strength of the minoritY vote.
t'Tf,e reason that a minority group making sueh a ehallenge must sltow,

as a threaholtl matter, that It is sullleiently large and geographieally

eompact to eonetitute a majorlty in a single-member district is this: Unless

mlnority voters possess Lhe potentiol to elect representatives in. the

absence of the chaltengetl structure or praetiee, they eannot claim to have

been lnJurett by that strueture or practice. The single-member tlistrict is

generaiy the appropriate standard against which to measure minority

froup potential Lo eleet because it ls the smallest political unit from which

iepresentatives are eteeted. Thus, if the minority group is spread evenly

throughout a multimember distriet, or if, although geographically eompact,

the mlnority group le eo smalt in relation to the surrotrnding white popula-

tlon that lt could not constitute a maJorlty ln a single-member district,

these mlnorlty votera eannot maintain that they wottld have been able to

eiect representativee of their choice ln the absence of the multimember

eleetoral strueture. As two commentators have explained,
,,To demonstrate [that minorlty voters are inJured by at-large electionsJ,

the minority voteE must be sulnciently coneentrated and politically



,.rc .r r., ).,r...;-..-.-r-.lilllii'i.Ji.aix,strrlEal{tltr{$d*ri-+r ' :T

18 THORNBURG u. GINGLbS

su.prq at 616. See atso, Blacksher & Menefee 6l-56, 68;

Bonapfel 366; Carpeneti 696; Davidson 4; Jewell 130' Q:q-
ond, ihe minority group must be able to show that it is politi-

eally eohesive. ti ttre minority gtgrp is not politieally.eo-

h"rir", it eannot be sald that the selection of a trlultimembef

eteetoral strdcture thrfarts dlstlnetive tninority group ln-

m*.tt. Blaeksher & itenefee 6l-65, 68-60, end tr' 344;

d"ip"r"ti 696-6fI; Davidson 4. Thitd, the minority tnust be

able to demonstrate that the white majority votes suf6eiently

as a bloe to enable it-in the abbence of speelal eireum-

rtn*Jt, such ad the tnlnorlty eandidate funning ilnopposed,

aen, inf,ra,et ---tisually to defeat the minorltV's prei9rl1d
eandtdate. See, e.9., Blaekshef & Menefee 611 53, b6-6?,

oo. ;Cf. Rogni, sufrfr, at 616-61?; Whitnomb, Eupry, ?\
iri-trs; UiUiti"" v. bscamhia cwntg, lla.r t!a.F' 2d

fOiZ, friae (CA6 1984). In establishing this lBBt eireum-

,[Jn"", the rnirroritg group deinonstrates that subrnbrgenee

in a white trttrltimember aistriet tmpedes its abtlity to bleet itS

ehoseh representatlves.-. 
Fiiliy: rte observe that the usual pred-letability of the

rnajotity,s sueeess distinguishes struetural dilution from the

meie loss of an oceasional eleetion. Cf. Daris v. Band,emtr,

- 
U. S. 

-, - 
(1986) (oplnion of Wrnto, l.); Bold'en,

,up*, at lll, n. ? (MAnsIrALL, J., dissenting);Whit'comb,
t r,p*, at 163. See also Blaeksher & Menefee 67, n' 333;

Note,'Geometry and Geography: Raeial.Gerrytnanderln-g 
-and

itr" Voting nignts Aet, 94 Yaie L. J' 189, 200, n' 66 (1984)

(hereinafter Note, Geometry and Geography)'

eoheslve that a putatlve dlatrletlng plen would reeult in dlstrlets ln whleh

m"mb"o of a raeial mlnorlty would eonstitute e mqiorlty o[ the-voters'

whose elear eleetoral eholces ere ln faet defeated by at-large votlng' 
. 
II

;i*6t ;;"rs, resldenees are rubstdntlally integrated throughout. the.Ju-

;;iJilr, it e at-targe dlrtrlet eannot be btamed fof t'he defeat of mlnorlty-

nrrntxrrted eandidates . . . . [Thle shndardl thus would only proteet raclal

;i#;6;;, iio. ai.lnu6on proxtmatety caused by rhe dterrterils plani

ic toould t4,l oscurr- runial tnitnriliea ptoportional rcpreac'nlntion.,'

ilU.f,"ft"t & ltenefee 65-88 (footnotee omitted) (emphesls added)'

,t

THORNBURG u. GINGLES l9

III
RACIALLY POLARIZED VOTING

Having stated the general legal prineiples relevant to-

elaims that $ 2 has been violated through the use of
inultimembet distriets, we turn to the arguments of appel-

lants and amicus curiae the Llnited States addressing

racially polarized voting.rE First we deseribe the District
Courtb tieatment of raeially polarized voting. Next, we

eorrsider appellants' claim that the Distriet Court used an

lneorreet legal standard to determine whether raeial bloc

toting in thJ eontested distriets was sufficiently severe to Ae

togniiable ds an element of a $ 2 claim. Finally, we consider

aplellants'eontention that the trial eourt employed an incor-

teet dehnitioh of faeially polarized voting and thus errone-

ously relied on statistical evidenee that was not probative of

raeial bloe voting. 
A

fHE DISTRICT COURT'S TREATMENT OF RACIALLY
POLARIZED VOTING

The investigation conducted by the District court into the

question of racial bloc voting eredited some testimony of

lay witnesses, but relied prineipally on statistical evidence

presented by appellees'expert witnesses, in particular that
iff"t"d by br.- bernard Grofman. Dr- Grofman collected

and evaluated data from 63 General Assembly primary and

general elections involving btaek candidacies' I'hese elec-

t-lon. *"." heltl over a period of three tlifferent election years

in the six originally ehallenged multimember tlistriets''' Dr'

Grofman subjeeted the data to two eomplementary methods

of analysis-extreme ease analysis and bivariate ecologieal

,rThe i,erms .,raeiatly polarized voting" and "raeial bloc vtrting" are usetl

lnterehangeably throughout this opinion.
,'The 1682 reapportionment plan left essentially undisturbetl the l97t

pton for flve of tfie original six contested multimember tlistricts. House

bi"trict 39 alone was slightly modifled. Brief for Appellees 8'



m THORNBURGU. GINCLES

regression analysis'-in order to determine whether blaeks

and whites in these distriets differed in their voting behavior.

These analytle teehniques fielded data eoneerning the voting
pa[terns oith" two iaees, lneludihg estimates bf the pereent-

ages of members of eaeh raee who voted for blaek eandidates.

The eourt's initial eoneideration of these data took the form

of a three-part inquiry: did the data reveal any eorrelatlon be-

tween the raee of the voter and the selection of eertain eandi-

dates; was the revealed eomelation statlstieally signifleant;

ahd was the differenee in blaek and white voting patterns

"substantively signiffeant?" The Distriet Court found that
blaeks and whites generally preferred different candidates

and, on that basis, found voting in the districts to be raelally

eotrelated.tt The eourt aeeepted Dr. Grofmart's expert opin-

ion that the correlatloh between the raee of the voter aird the

voteris ehoiee of.eertain eandidates was statistieally signlft-

eant.t Finally, adopting Dr. Grofman's terminology, see Tr.

'fire Dlstrlct court found both methods etandard ln the lltcrature for

the anatysis ofraeiatly potarlzed votlng' 690 F. Supp., et 36?-378,nn- 28,

and 32. See also Engstrom & MeDonald, Quantltative Evidence ln Vote

Ditutlon Lltigation: pollttcal Partleipatlon and Polarized Voting, l? Urban

Lawyer 369 iSummer 1985); Grofman, Migalskl & Novlello, The "Tot'allty
of Cireumstances Test, ln Sectlon 2 of the 1982 Extension of the Votlng

Rlghts Aet: A Soelal Sclence Perspeetlve, ? Law & Polley 199 (Apr' 1985)

(hereinafter Grofman, Illlgalskl, & Novlello).
lrTtre eourt used the teim,,raelal polarlzatlon" to deeerlbe thlo correle-

tlon. lt adopted Dr. Grofman'e deffnitlon-"raelal polailzatlon" exlste

where there li ,,a eonsletent relatlonahlp between [thel raee of the votet

"nJth"'rr"y 
ln whleh the voter Yotee,'TI. !60, or to p,t lt differently,

where'tlaek voters end white voters vote dlfferently," lil', al2p71' We,

too, atlopt this def,nltlon of "reelal bloe" or "reelally polerlzid" votlng.

See, irfro, at 

-.,The eourt found that the dete reffeeted positlve relationehlps end that

the eorr.etatlons dld not happen by ehenee. 690 F. Supp' 368, and n' 30'

See also D. Barnee C J. ionley, Statlatieal Evidenee ln Litlgatlon 32-M

itgg6); tr'i*h"t, Multlple Regtesslon ln tegal Proeee6ings, 80 Colum' L'
h"* Coz, 716-7?0 (1980); Grofman, Migalskl, & Noviello 206.

THORNIIURG u. GIN(ILES 2l

195, the eourt foulrd that in all but two of the Il3 electiotrs'''

the degree of racial bloc voting was "so marked as to be

substantively significant, in the sense that the results of the
individual eleetion would have been different depending upon

whether it had been hekl among only the white voters or only

the blaek voters." 690 F. Supp., at,368.
The court atso reported its findings, both itr tabulated nu-

merieal form and in written form, that a high percentage of
blaek voters regularly supported black eantlitlates and that
most white voters were extremely reluctant to vote for black

eandidates. The court then eonsideretl the relevatlce to the

existenee of legally significant rvhite bloc voting of the fact
that blaek eandidates have won some eleetiotls. It rleter-
mined that in most instanees, speeial eireumstances, sueh as

lneumbency and lack of opposition, rather than an a diminu-
tion in usually severe white bloe voting, accounted for these

eandidates' sueeess. The court also suggested that blaek

voters' relianee on bullet voting was a significant factor in
their sueeessful efforts to eleet eandidates of their choice.

Based on all of the evidenee before it, the trial court con-

cluded that eaeh of the distriets experiences raeially polarized

voting "in a persistent and sevete tlegree." 600 tr. Supp.,

at 367. 
B

THE DEGREE OF BLOC VOTING THAT IS LEGALI,Y
SIGNTFICANT UNDER $2

I
Appellants' Arguments

North Carolina and the United States argue tha(, the test

iised by the Distriet Oourt to determine whether voting
pattertrs in the disputed distriets are racially polarizetl to an

extent eognizable uhder $ 2 will lead to results that are ineon-

sistent with eongressional intent. North Carolina maintains

aThe two exeeptions were the 1982 state House electiorrs in l)istricts 2l

and 23. 690 F. Supp., at 368, n. 31.



I

TH0RNBURG u. GINGLES

that the eourt eonsidered legally signifieant raeially polariZed
voting to oeeur whenever "less than 60% of the white voters
east a ballot for the black eandidate." Brlef for Appellants
36. Appellants also argue that taeially polarized voting is

legally iignifteant only when it always results in the defeat of
black eandidates. 1d,., al39-40.

The United States, on the other hand, isolates a slngle line
ln the eourt's oplnlon and ldentiflee it as the eourt's eotnplete
test. Aeeording to the Solleitor General, the Distrlet Court
adopted a standard under which legally signiffeant raeial bloc
voting ls deemed to exlst whenever "'the results of the
individual etectiofi woutd have been different depending upon
*hethet tt had beetr held emong only the white voterd or ohly
the blaek voterc ih the eleetioh.l " Brief for United States aB

Amtan Cilria,e 29 (quoting Ginlles, Sm F. Supp., at $68).

We read the Distrlet Court opinion differently.

2

Thn Stnnd.ard,for Legahy Significant Radal BlocVotin4

The Senate Report states that the "extent to whieh votlng
in the eleetions of the state or politieal suMivision ls raelally
polarized," S. Rep. 29, is relevant to a vote dilutlon elaim.

Further, courts and eommentatofs agree that raeial bloe

voting is a key eleinent of ri vote dilution elaim. .See, e. 9.,
bscainbia County, ?48 F. 2d, at 1043; United States v.

Mamngo County Conm'n,?3l F. 2d 1546, 1666 (CAll 1984);

Neaett-v. Sidns,6?l F. %lm9,22S(JC,A6l9?8), eert. deirled,

4dG U. S. gff (1980); Johnson v. Halifan County, 694

F. Supp. 161, 1?0 (EDNC t984); Blaeksher & Menefee;

nngstrom & Wldgen, Piuning Thorns from the Thieket: Ah
Empirieal Test for the Exlstence of Raeial Gerrymandering-,
Z Lesls. Stud. Q. 4or, log (19??) (heieafter bnstrom &

frynagen); Parker 10?; Note, Geotrtetry and Geography l!9.
Beeause, ds we explain bblow, the extent of bloe voting
heeessary to demonstrate that a tninofity's ability to elect its
preferred reprebentitlves is impaired vailes decoidlng to

THORNBURG u. GINGI,FIS

several faetual eireumstanees, the degrce of bloc voting
which eonstitutes the threshold of legal significance will vary
from district to district. Nonetheless, it is possible to state
Bome general prineiples and we proeeed to do so.

The purpose of inquiring into the existenee of raeially po-

larized voting is twofold: to aseertain whether minority group
rnembers eonstitute a politically eohesive unit and to deter-
mine whether whites vote sufffciently as a bloe usually to de-
feat the minority's preferred candidates. See, szpro, at

Thus, the question whether a given district experi-
enees legally signifieant raeially polarized voting requires dis-
erete inquiries into minority and white voting praetiees. A
showing that a signiflcant number of minority group mem-

bers usually vote for the same eandidates is one way of prov-
ing the political eohesiveness neeessary to a vote dilution
elaim, Blaeksher & Menefee 69-60, and n. 344, and, conse-

quently, establishes minority bloe voting within the context
of $2. And, in general, a white bloe vote that nornrally will
defeat the combined strength of minority support plus white
"erossover" votes rises to the level of legally signifieant white
bloe voting. Id., at 60. The amount of white bloc voting
that ean generally "minimize or eancel," S. Rep. 28; Regester,
412 U. S., at 766, blaek voters' ability to eleet represent-
atlves of their ehoice, however; will vary from district to dis-
trict aecording to a number offaetors, including the nature of
the allegedly dilutive electoral mechanism; the presenee or
absenee ofother potentially dilutive electoral deviees, sueh as

tnajority vote requirements, designated posts, and prohi-
bitions against bullet voting; the pereentage of registered
voters ln the distriet who are members of the minority gl'oup;

the size of the distriet; and, in multimember districts, the
huinber of seats open and the number of canditlates in the
fietd." See, e. g., Butler 8?4-8?6; Davidson 5; Jones, The

lmpact of Loeal Election Systems on Black Political Repre-

'This llst of feeiors is lllustratlve, not eomprehensive.



24

sentation, ll Urb. Aff. Q. 346 (19?6); U. S. Cotntnissiofl on

Ctvil Rlghts, The Voting Rlghts Ach Untulfilied Goals 38-41
(1e81).' 

Beedtise loss oi potitieai power throttgh vote dilution ls dis-
tinet frotn the mere tnability to win e partieular electloh,

' 
Whitnomb, 403 U. S.; at 163, d pattern of racial bloe voting
that extends over a period oi tlnte is more probative of e
elaim that e distilet experienees legally signifleant polariza-

tion than ate the lesults of e single election.t Blaeksher

& Menefee 61; Note, Gebmetry and Geography 200, h. 66

("Racial polaiizatioh should be seen as att ettribute not of e
single eleetion, but rather of a polity viewed over time. The

.on."rn is heeessarily temporal and the dnalysis historieal
beeause the evil to be avolded is the subordination bf ininor-
ity groups in Amerlean politics, not the defedt of irrdlvlduals

in partteutar eleetoral eontests"). Also for this reason, in a
disirict where eteetions are shown usually to be polarlzed,

the faet that raeially polarized voting is hot present in one or
a few individual eleetions does not neeessarily negate the eon-

eluslon that the distriet expefiences legally slgnifieent bloe

voting. Furtheimore, the sii0cess of d rrtlnority cdndldate ih
e pai,icrrlar electloh does not heeessarily prove that the

distrtet did not experlenee polailzed voting ln that eleetion;

speeial elreumstanees, sueh es the absbnee of ah opponent,

lncutnbehey, or the utillzation of btrllet vo[ing, rnay explairt

mihority eleetoral BueeeBB ln a polarized eontest.a

THORNBURG u. GINGLbS

EI'he number of eleetlone thai muat be etudled ln order to determine

whether votlng ls polarlzid wlll vary aeeording to pertlnent elreumstenees.

One lmportaniclreumetanee ls the number of eieetlons ln q,hleh the tnlnor-

liy grorp has sponsored cendldates. Where d mlnorlty qrouq h.as n1ye1

b""i 
"bi" 

to sponsoh a eandidete, courte must fely on other faetors that

tend to prove uneqUel aecess to the eleeboral proeess. Slmllarly, where a

minorlty group has begun to sponsor eandidates Just reeently, the faet thet

statlstlcs"from otrty one or e few eleetlons are available for exatnlnation

doea no[ foreelose u vote dilutlon elalm.
iTtrla tlat of epeelal elriumstrnces ls lllustratlve, not exetlslve'

THORNBURG o. GINGLES

As must be apparent, the degree of racial bloc voting that
is cognizable as an element of a $ 2 voLe dilution claim will
vary aeeording to e variety offactual eireurnstances. Conse-
quehtly, there is no simple doetrinal test for the existence of
legally signilicant raeial bloe voting. llowever, the forego-
ing general prineiples should provide eourts with substantial
guidanee in determining whether evidenee that black and

white voters generally prefer different candidates rises to the
level oflegal signifteanee under $2.

3

Standard Utilized ba thn Dietrict Cou.rt

The Distriet Court clearly did not employ the simplistic
standartl identified by North Carolina-legally significant
bloc voting oeeurs whenever less than 60Vo of the white vot-
ers east a ballot for the blaek eandidate. Ilrief for Appel-
lants 36. Arrd, although the District Court did utilize the
measure of "substantive signifleanee" that the United States
aseribes to it-"the results of the individual eleetion would
have been different depending on whether it had been held

among only the white voters or only the black voters,"' Brief
foi United States as Amicus Curia'e 29 (quoting Gingles,690
F. Supp., at 368)-the eourt did not reaeh its ultimate eonelu-

sioh that the degree ofracial bloc voting present in each dis-

triet ls legatty significant, through mechanieal relianee on this
standard.' While the court did not phrase the standarcl for
legalty slgnifieant racial bloc voting exaetly as we do, a fair
reading of the eourt's opinion reveals that the eourt's analysis

eonforms to our view of the proper legal standard.

"The trlal court dld not aetually employ the term "legally signiflcant."

At times lt seems to have ueed "substantive signifieanee" as Dr. Grofman

dld, to deeerih polarlzation severe enough to result in the eeleetion of
dillerent eandidatee ln raeially separate electorates. At other tlmes,

however, the eourt used the term "eubstsntively signifieant" to refer to its

ulilmate dctermination that raeially polarized voting in these districts is

sulliclently severe to be relevant to a $ 2 clairn.



X3 THORNBURGU. GTNGLES

The Distriet Court's findings eoneerning blaek support for
btaek eandidates in the flve multimember distriets at issue

here elearly establish the politieal cohesiveness of blaek vot-
ers. As is apparent from ihe Distriet Court's tabuiated ffnd-
ings, reprodueed in Appendlx A, infra, blaek voters'support
for blaek eandidates was overwhelming in almoet every elee-

tion. In all but 6 of 16 primary eleetions, black support for
blaek candidates ranged between TlVo and 92%; and in the
general eleetions, blaek support for blaek Democrat candi-
dates ranged between 87%,and96%.

ln sharf contrast to its ffndings of strong blaek support for
blaek eandidates, the Distriet Court foutrd that a substantial
tnajority of white votels would rarely, lf ever, vote ior a

btaek eandidate. trn the primary elections, white support for
black eandidates ranged betweeh 8Vo and 60Vo, and ln the
general elections lt ranged between ?8% and 49%. See

Lppendix A, in\a. llhe eourt also determlned. that, on

"u"r"g", 
Sl.7% rf whit" voters did hot vote fof any black catr-

didatJ in the primary eleetlons. In the general blectlons,
white votera almost always lanked blaek eandidates either
last or neit to last in the multieandidate ffeld, except ln
heavity Demoeratie areas shere white voierg eonsistently
ranked black candidates last among the Demoerats, if trot last
or next to last among all eandidates. The eourt further ob-

served that approximately two-thirds of white voters did not
vote for blaek- eandidates in generel eleetioits, even after the
eandidate had won the Demoeratle primary and the ehoiee

was to vote fol a Republican or for ho one.a

THORNBURG U. GINGLES

While the Distriet Court did not state expressly that the
percentage of whites who refused to vote for black candidates
in the eontested districts would, in the usual eourse of events,
result in the defeat of the minority's candidates, that conclu-
sion is apparent both from the eourt's faetual findings and

from the rest of its analysis. First, with the exception of
House Distriet 23, see infra, aL-, the trial court's ffndings

elearly show that blaek voters have eqioyed only minimal and

sporadic suceess in electing representatives of their choice'
See Appendix B, infra. Second, where black eandidates
won eleetion, the court closely examined the eireumstances of
those elections before eoncluding that the 8ucee88 of these

blacks did not negate other evidence, derived from all of the
elections studied in eaeh district, that legally signifieant ra-

eially polarized voting exists in eaeh district. For example,

the eourt took aeeount of the benefits incumbeney and run-
hing essentially unopposed eonferred on some ofthe sueeess-

ful blaek eandidates,o as well as of the very different order of

however, lt ls elear from the trial court'a tebulated ffndings and from the

exhlblts that were before lt, I App., Exs. 2-10, that the court relied on

data that was epeeille to eaeh individual dietrict in eoncluding that each dis-

trlet experleneed tegally slgniflcant raelally polarized voting.
rFor exampte, the eourt found that lncumbeney aided a suceessful blaek

eandldate ln the l9?8 primary in Senate Dlstrlct 22' The eourt also noted

that ln House Distrlet 23, a blaek eandidete who galned eleetlon in l9(8,
1980, end 1982, ran uneonteated ln the l9|l8 general eleetion and in both the

prlmary and generat electlons ln 1980. In 1982 there was no Republican

opposition, a faet the trial eourt interpreted to mean that the general

eiectlon was for alt practlcal purposes unopposed. Moreover, in the 1982

Drlmary, there were only two white eandidates for three eeats, so that one

Ltaek candldate had to suceeeti' Even under thie eonditlon, the court

iemarked, 68% of white voters still refused to vote for the blaek lncum-

bent-who was the eholee of 90% of the bleeks. In House Distrlet 2l'
where e blnck won election to the alx-member delegatlon in l9B0 and 1982,

the eourt lound that ln the relevant prlmarles approximately 6o% to70% of
whlte votefe dld nol vote for the black candldate, whereas approximately
$96 of btaeks did. The eourt additlonally observed that although wlnning

the Democraile primary ln this district ls historieally tantamount to

. ln statlng thrt 81.?% bf whlte voters did hot vote for rhy blach candl-

dates ln the"brimery electlon end that twri-thlrds of whlte Vo[ers dld no[

vote for black candldater ln general eleetions, the Dlstrlct CoUrt aggre-

cet€d deta from atl ek ehallenged multlmember dlstrleta, apparently for

iaae of reportlng. T1e lnqulry lnto the exlstenee of votc dllutlon eauged

bi rubmeipnee ld a multlinembbr dlstrlet ls dlstrlct-speclf,e. When eon-

aidetlng severat separate vote dllutlon ctalms ln a sinde ease, eourta muet

notrclj ontlate aggregstad from ell thd ehallenged diatrlcts lh cotteludlng

ttut tailally poterl-zed votlng erlah lrt eaeh distrle[. ln i.he lnateht ease,



% THORNBUBG u. GINGLES

preferenee blaeks and whites assigned blaek eandidates,'in
ieaehing its eonelusion that legally signifleaht raelal polarlza-

tion exists ln eaeh distrlet.
W" .on.lude that the Distriet Corlrtis appioaeh, whieh

tested data derived from thtee electioh years ln eaeh distrlct,
and whieh revealed that blaeks strongly supported blaek ean-

didates, whlle, to the blaek cendldates' usual dettimeht,
whites rarely did, satisfeetorily addressea eaeh faeet of the

Proper legal standard. 
C

EVIDENCE OF ITACIAT,TY POLARIZED VOTING
I

Appellnils' Aryurinnnt

North Carolina and the United States also eontest the evi-

denee rlpon $hieh the Distriet Court relied in finding that
votlng patterns in the ehallenged distriets were raeially Po:
larized. They argue thet the tern "rdelally polarized voting"
inust, ris a mattei of law, tefer to voting pattef! fq -*.t":l
Lhe prinnipal ca&e ls taee. They eorrtend that the Dtstrlet
Cou* utilized d legally ineorreet deffnition of reelally rytg-
ized voting by relying on bivariate statistieal analyses whlch

merely de]monstrated a conelat'ionbetween the raee of the

vohrlnd the level of voter sUpport for eertaln candidates,

but whieh did not prove that raee was the prlmary determl-
nant of voters' ehliees. Aeeording to appellants and the

United States, only multiple regression analysis, whieh can

take aecount of other variables whleh might also explain vot-
err'ehoiees, sueh as "party afflliation, age, religion, ineome[,J

eteetlon, 66% of whltes deetlned to vote for the Demoeratle blaek eandldate

in the general electlon.
,ItJ eourt noted thet ln the 1982 prlmary held in House Distrlet 36, out

of" n*tA ofeight, the sueeessful btaek eandidate was rlnked f,ret by blaek

Voters, but tivetrth by whlteo. Slmltdtly, the court fo[nd that the two

rrr"t . *t o won seats ln the ffve-member delegatlon from House Dlstrlet 89

were ranked first and second by,blaek voters, but seventh end elghth by

whlte voters.

TUORNBURG u. GINGLES 29

ineumbency, education, eampaign expenditures," Brief for

Appellants i2, "media use measured by eost, . . . name, iden-

tifleation, or di*t"n." that a candidate lived from a particular
precinet,;' Brief for United States aa Amicus Curiae 30,

n. 6?, ."n prove that race was the primary determinant of

voter behavior.tt
Whether appellants and the United States believe that it is

the voter's ra." or the eandidate's raee that must be the pri-
inary determinant of the voter's choiee is unclear; indeed,

theii eatalogs of relevant variables suggest both.D Age, reli-
gion, income, and edueation seem most relevant to the voter;

Incumbeney, campaign expenditures' name identifieation,
and mbdia use are peitinent to the eandidate; and party affili-
atlon eould refer 6oth to the voter and the eandidate' In
either ease, we disagree: For purposes of $2, the legal con-

eept of raeially polaiized voting incorporates neither causa-

tion nor intent. It means simply that the raee of voters eor-

felates with the seleetion of a certain candidate or candidatee;

that is, it refere to the situation where different races (or

fninorily language groups) vote in blocs for different candi-

dates. 
-Grofman, 

Migalski & Noviello 203. As we demon-

strate below, appellants' theory of racially polarized voting

would thwarl dre goals Congress sought to aehieve when it
amended $ 2 and would prevent courts from performing the

"funetlonal" anatysis of the political proeess' S' Rep' 30,

h. 119, and the ;searehing praetieal evaluation of the 'past

D Appellants argue that plalntilfs must establish that raee was the

prlmary determlni'nt of voter behavior as part of their primalocde showing

orpot"rr.a votlng; the solieitor General suggesta that plaintilfs make out a

iima facie eaeeherety by showing a goyelatign between race and the
'aeleetion 

of eertaln eandidales, but that defendants should be able to rebut

by showlng that faetors other than r&ce were the prineipal eausee of voter€'

eholees. We rejeet both argumentr.
lThe Fifth Cireuit eaEea on whleh North Carolina and the Solieitor

Geiierat rely for thelr posltlon are equally ambiguoua. see Lee County

Brurch o1 t't A,lC e v. C ity of Opetiko, 7 48 F . 2d 1473, f 482 ( I 984); J on'es v'

Cily ofiubbock,TS}F. 2d 233,234 (1984) (Hlgginbotham, J', concurring)'



SO THORNBURG U. GINGLES

afld present feality,"' id.t aL80 (footnote omltted); mandated
by the Senate Report. 

z

Caucation l,releuan| tn Seclion I In4uirg
The flrst reason se reject eppeiiants' argument that ra-

eially polarized voting refers to voting patterns that are ln
soml way catned,by mne, rather than to votlng patterns that
are merely conelated, u,ith tltz mce of th,e ooler, is that the
reasons blaek and white voters vote differently have no rele-
vanee to the eentral tnquiry of $2. By contrast, the correla-
tlon butwtien raee of voter and the seleetlon of eertaln cdndi-
dates is eruelal to that inquirY.

Both $ 2 itself and the Senate Report make clear that the
eritieal questlon ln a $2 elaim ls whether the use of d eon-
'tested elietoral praetlee or stnreture fesults ln members of a
proteeted group havlng less opportunlty than other rnembers
of the eleetorate to partleipate in the politleal process and to
eleet representatives of their eholee. See, e. 9., S, Rep. 2,

X, %,29, n. 118, 36. As we explained, sttpt'o, at 

-,multimember distrlets may lmpair the ability of blacks to
eleet representatlves of their ehoiee where blaeks vote suffl-
eiently as a btoe as to be able to elect thelr preferred candi-

dates irt a blaek mqfority, single-member dlstriet and where a
white mqiority voies sufnciently as a bloe usually to defeat
the eandi.-datei ehosen by blacks. It Is the differenne b*
tween the eholces made by blaeks and whltes-not the rea-

sons for that difference-ihat results ln blaeks havlng less

opportunity than whltes to eleet their preferted 
^reptesent-

"[ir.r. Cinsequently, we eonclude that under the "res':lts
test" of $ 2, onlythe eorrelation between raee of voter and se-

leetlon of 
"""t"in 

eandidates, not the eauses of the eofrele-
tion, matters.

Tire lrrelevahee to n $ 2 tnquiry of the feasohs why blaek

dnd whlte voters vote diffetetrtly suppotts, by ltself, our

reJeetion of appellantsr theory of raeially poletlzed votlng'
H6wever, their theory eontains other equally seflotrs flaws

THORNBURG U. GINGLES

that tnerlt further attention. As we demonstrate below, the
addition of irrelevant variables distorts the equation and

fields results that are indisputably ineorrect under $2 and
the Senate Report. 

B

Race of Voter as Primary Detenninnnt of Voter Behnttior

Appellants and the United States eontend that the legal
eoneept of "raeially polarized voting" refers not to voting
patterns that are merely conelated, uith thp aoter's race, but
to voting patterns that are d'etennined, primnrilg by the

aoter's race, tather than by the voter's other soeioeeonomie

eharaeterlsties.
The ffrst problem with this argument is that it ignores the

feet that members of geographieally insular racial and ethnic
groups frequently share soeioeeonomie charaeteristics, such

as income level, employment status, amount of education,
housing and other living conditions, religion, language, and

so forth. See, e. g., Butler 902 (Minority group "memberE'
shared eoneerns, ineluding political ones, are . . . a function
of group status, and as sueh are largely involuntary. . . . As a
group blaeks are eoneerned, for example, with police brutal-
ity, substandard housing, unemployment, ete., because these
problems fall disproportionately upon the group"); S. Verba
& N. Nie, Partleipation in America 161-162 (1972) (hereinaf-

tei Verba & Nie) ("soeioeconomic status . . . is closely re-
lated to faee. Blaeks ln Ameriean society are likely to be in
lower-status Jobs than whites, to have less edueation, and to
have lower ineomes.") Where such eharacteristics are
shared, raee or ethnic group not only denotes color or plaee of
origin, it also funetions as a shorthand notation for common

soeial and eeonomie characteristies. Appellants' definition of
raeially polarized voting is even more pernieious where
shaied eharacteristies are eausally related to raee or ethnic-
Ity. The opportunity to achieve high employment status and
ineome, for example, is often lnfluenced by the presenee or
absenee of raeial or ethnie discrimination. A definition of ra-



r&b'.&*r.....iE.

82 THORNBURGU. GINGLES

elally polarized voting *i,i.h holds that blaek bloe voting does

not exist when blaek voters' ehoice of certain eandidates is

most strongiy Influeneed by the fact that the voters heve low

Ineomes and tnenlal Jobs -when 
the feason most of those

voters have menial jobs and low ineomes is attributable to
past or preseht raeial discrlminatloti-mns eounter to the

Senate lieport's lnstruction to eonduet a searehing dhd prdc-

tleal evaluatioh of past and present reality, S. Rep. 30, and

lnterferes with the purpose of the Votitrg Rights Act to eliml-

nate the negative effeets of past diserlmination oh the elee-

toral opportunities of mlnorities. 1d,., al6, 40.

Furthermore, under appellants' theory of raeially polarized

voting, even uneontrovertlble evldenee that eandldates

strongly preferred by bleek votert are always defeated by a
bloe voting white maJority would be dismissed for failure to
pnove raeial polarlzation whenever the blaek and white popu-

iations eould be described ln terms of other soeioeconomic

*ififi:frlii, 
""rr." 

i raeially-tnixed, urbah multlmetnber

distriet ln whleh blaehs end whltes po88e88 the aame eoeioeeo-

noinle eherieteiistles that the reeord ln thls case ettrlbutes to

blaeks and whites ln Halifax County, a part of Senate Distriet
2. The atrhual mean lheome for blaeks ln this dlettlct le

310.466. end 47.8% of the blaek eotnmunlty llves in poverty.

iioi" than halfjl.6%-of bleek adults ovei the age of 26

have only an eighth grade edueation or less. Just over half

of blaek-eitizeiri reside irt their own homes; 48.9% live irl
rental uhits. And, almost a third of all black households are

without a eai. In eontrest, onty l2.g% of the whites in the

distriet,live belort the poverty line. Whites enjoy a mean in-

eome of $19,042. White residents are better educated than

, blaeks---only 26.6% of whites over the age of 26 have only an

elghth grade edueation or less. Furthermore, only 26'2%.9f

wfrii"riir. iir rentel units, arrd only lO.Zf tlv6 tn hortseholds

ivith rlo vehlele evailable. I ApF., Ex-44. As ls the ease ltt

THORNBURG u. GINGLES 3:l

Senate District 2, blaeks in this hypothetical urban district
have never been able to eleet a representative oftheir choice.

Aeeording to appellants'theory of racially polarized voting,
proof that blaek and white voters in this hypothetieal district
iegularly ehoose different candidates and that the blacks'
preferred candidates regularly lose eould be rejected as not
probative of raeial bloc voting. The basis for the rejection
would be that btacks ehose a certain candidate, not princi-
pally beeause of their race, but prineipally beeause this eandi-

date best represented the interesta ofresidents who, beeause

of their low incomes, are partieularly interested in govern-
ment subsidized health and welfare serviees; who are gener-

ally poorly etlucated, and thus share an interest in job tryln-
ing programs; who are, to a greater extent than the white
eommunity, eoneerned with rent eontrol issues; and who

favor major public transportation expenditures. Similarly,
whites would be found to have voted for a different candi-

date, not prineipally beeause of their raee, but primarily
because that candidate best represented the interests of
tesidents who, due to their edueatlon and ineome levels, and

to their property and vehiele ownership, favor gentriflcation,
low residenlial property taxes, and extensive expenditures
for street and highway improvements.

Congress eould not have lntended that eourts employ this

deftnition of racial bloe voting. First, this deftnition leads to

results that are ineonsistent with the effects test adopted by

Congress when it amended $ 2 ahd with the Senate Report's

admonition that eourts take a "funetional" view of the politi-

eal proeess, S. Rep. 30, n. 119, anrl eonduct a searching and

practieal evalttation of reality. /d., at 30. A test for racially

polarized voting that denies the faet that raee and soeioeeo-

nomie eharaeteristies are often elosely eorrelated permits nei-

ther a praetieal evaluation of reality nor a funetional analysis

of vote dilution. And, eontrary to Congress' intent in adopt-

ing the "results test," appellants'proposed deftnition could

result in the inabitity of minority votetrs to establish a critieal



si THORNBURGU. GINGLES

eiemeht of a votd dilutlotr elalm, even though both raeee

engage in "monolithie" bloe voting, td,., at83, and genefatlons

of bhek voters have been unable to eleet e representatlve of
thelr ehoice

Secohd, appellatrts' tntcrpietatloh of "raelally poletized
Votlng" ereeles ah lrieeoneilable tensiotr between thelr pp-
poserl'tteatmeht of soeloecohomle eharaeterlstieo ln the bloc
Votitrg eohtext *nd the Sendte Report's statement that "the
exten-[, to whieh members of the minority group . . . bear the
effeete of dlsetiminatlon in sueh sreas as edueation, employ-

ment ahti health" iney be nilevant to d !2 elaim. Id,.; at29.
lVe eah ffnd no support ln elther logle or the leglslative hls-
tory for the anomalous eonelusiotr to whlch appellants'posl-
tiotr leeds-that Congress intended, oh the one hand, that
proof that a mlnotlty group ls predomlnately poor, unedrt-

.at"d, and unheelthy should be eonsldered e faetof tendingto
iriove d $X vtolatlon; but thet Congress lntehded, oti the
other hand, that pfooi that the same boelooionomle ehafac-

terlstles greatly lnf,uenee blsck voters' ehblee of eahdidates

should destrolthese voters' ability to establlsh one of the
most lmpoitatit elements of e vote dtlutloir elalrn.

n

hace of Cand,iilate a Pdnmg Detenninmtt of
Votcr Behador

North Carollna's end the Untted State's euggestlon that
raeially polarized votlng means that.voterd seleet or reJeet

eatrdldetis principtly on the basle oi lhe eanl,idnle'a rtir,e la

"fr*fffif;l'n","r*"*" of !2 ahd e funetlonel understeird-
lng of t'he phenomenon of vote dilutlon mahdate the eonelu-

slJn that the raee of the eendldate p er sels lrreletant to raelal

bloe voting anelyeis. seetion 2(b) statea that a vlolntion ls

"rt"tti.tr"i 
tf ltian l,6 shown thet members of e pnoteeted

mlnorlty group "have lesc opportunlty then other membere of
the eleeto"rate t, . . . eleet-representatlies of ttwir chnine'"

THORNBURG U. GINGLES

(Emphasis added). Beeause both minority and m4jority vot-

ers often select members of their own raee as their preferred

representatives, it will frequently be the ease that a blaek

."ndid"t" ls the eholce of blaeks, while a white candidate is

the ehoiee of whites. Cf. Letter to the Editor from Chandler
Davldsotr, 1? New Perspectives 38 (Fall 1986). Indeed, the

faets of thls ease illustrate that tendency-blaeks preferred

blaek edndidates, whites preferred white candidates. Thus,

as a tnatter of eonvenienee, we and the Distriet Court may

refer tri the prefeired representative of blaek voters as the

"black candidate" and to the preferred representative of
white voters as the "white eandidate." Nonetheless, the fact
that race of voter dud raee of candidate is often eorrelated is

hot direetly pertinent to a $ 2 inquiry. Under $ 2, it is the

status of the eandidate as the chosen represenlntiae of a
particular xicial grmP, not the raee of the candidate, that is

'TTTli;rstanding of how vote dilution through submer-

genee in a white majoilty works leads to the same conelusion.

The essenee of a submergence claim is that minority group

tnerhberu pfefet certain candidates whom they eould elect

were lt not for the interaction ofthe challenged electoral law

or strueture with a white majority that votes as a significant

bloe for different eandidates. Thus, as we explained in Part
lll, suptt, the existence of raeial bloe voting is relevant to-a

vote diiution elaim in two ways. Bloc voting by blaeks tends

to prove that the blaelt community is politieally cohesive, that
is,it shows that blaeks prefer certain candidates whom they
eould eleet ln d single-member, blaek mqjority districh Bloc

votlhg by e white mqiority tends to prove that blaeks will
gun"rally be irnable to eleet representatives of their choice.

Clearly, only the raee ofthe voter, not the raee ofthe candi-

date, is ielevaht to vote dilution analysis. See, e. g',

Blaeltsher & Menefee 69-60; Giofman, Should Represent-

atives be Typleal? 98; Note, Geometry and Geography 2trI.



rlltrlri:ir*}....:r,fi.l,*r.r,o!.sergtl*trtl*lttIaaarg}s@l'&,ii.:'t,lt.'..l,l:J:

THORNBURG u. GTNGLES

Seeond, appellants' Buggestion that raeially polarized vot-
ing refers to votlng patterns where whites vote for white
.andid"ter beeause ihey ptefer ineinbers of thelr own raee or
aie hostile to bleeks, es opposed to votlng pattefns wheie
whites votd ftir white eandidates beeause the white eandi-

. dates spent more on their eampaigns, utllized more medla
eoverage, and thus eqfoyed greater hame fecognitlon than
the blaek eandidates, fails for ahothbf, lndependent redsoh.
This argument, like the argtrmeht that the raee of the voter
must 6 the primary determinant of the voter's .baltot, ls

ineonsistent with thb puiposes of ! 2 end would rehder tnean-

lngless the Senate Report feetor that addresses.theimpae!.9f
loi soeioeeonomie status on a minority group'e level of politl-
*t[H:'J'11t1;ded 

ura*he voting Rights Aet eradieate in-
equalities in politteal opportrtnities that exist due to the vesti-
gial effeets oipast purposeful diserimination. S. Rep. 6, 401

tt. n. Rep. No. W-\n, p. Sl (1981). Both this Court atrd

other federal eourts have reeognized that irolitical partleipa-

tion by ihinofitles tends to be depressed where mlnotity
group membets stiffer effeets of priof dlserlmlhatloh sdeh as

lnferior edueatioti, poor employtrreht opportunlties, and low
ineoines. See, e.-g., Whiln v. Regester, 412 U. S., at

?68-?69; Kirkaey v.-fimd, of Supentisorg of Hints Countg,
llisa.,664 F. 2d 139, 146-146 (CA6) (en bane), cert. denied,

4sd U. S. 96tl (19??). See also verba & Nle 102. the Seh-

ate heport [ekiiowledges thls tehdehey end instruets that
"the extent to whleh members of the mlnorlty group . . . beaf

the effeets of diserirnlnatlon ln sueh erees as edueatlon, em-

ployment and health, whieh hlnder theh abllity to partleipate

effeetively ln the politleal process," S. Rep. 29 -(footnote
omitted), is a faetof whleh may be probatlve of unequal

opportuntty to partieipate ln the politlcal pioeess atrd o eleet

r"i""r"rttatir"s. Courts ahd eommentators have recognized

irtth"t that eandldates geherally must spend mord monei ln

ordef to wln eleetiori ln e multinrerht oi dlstrlet than lrr e

THORNBURG U. GINGLES

single-member district. See, e. g., Graaes v. Barnns, 343 F.
Supp. ?04, 720-721(WD Tex. 1972), aff'd in part, rev'd in
part snb nnm. Whitp v. Regester, supna. Berry & Dye 88;

DaVidsoh & Fragd, Nonpartisan Slating Groups in an At-
Large Setting, in Minority Vote Dilution lD-123: Derfrrer
654, tr. 126;Jewell l3l; Karnig, Blaek Representation on City
Couneils, 12 Urb. Aff. Q. 22g,An (1976). If, because of in-
ierlor edueation and poor employment opportunities, blacks
earn less than whites, they will trot be able to provide the
eandidates of their ehoiee with the same level of financial sup
port that whites can provide theirs. Thus, eleetoral losses

by eandidates preferred by the black eommunity may well be

attributable in part to the fact that their white opponents
outspent them. But, the fact is that, in this instanee, the
eeonomie effeets of prior discriminatlon have combined with
the multimembei electoral strueture to afford blacks less

opportunity than whites to partieipate in the politieal proeess

and to eleet representatives of their ehoiee. It would be

both anomalous and inconsistent wlth eongressional intent to
hold that, on the one hand, the effects of past discrimination
whleh hlndei blaeks' ability to participate in the politieal
proeesa tend to prove a $ 2 violation, while holding on the
other hand that, where these same effects of past discrimina-
tloti deter whites from voting for blacks, blaeks eannot make

otlt a erucial element of a vote dilutiorr claim. Aeeord,
Escambia County,748 F. 2d, at.1043 ("'the failure of the
biaeks to solieit white votes may be caused by the effects of
past diserimlhation"') (quoting United States v. Dallas
counly cotnm'n, ?39 F. 2d 1629,1636 (cAll 1984)); Unitnd
States v. Iliwengo Cruntg Cotnm'n,73l F- 2d 1&16' 1567

(CAtl 1984), eppeal dismissed, eert. denied, 469 U. S. 

-(1e84). 
6

hacia,t Animnsity as Primary Determinant of
Voter Behauim

fi'lnally, we reJeet the suggestion that raeially polarized



THORNBURG U. GINGLES

vdtirrg refers only to whlte bloe voting whieh is eaused by

*hlt"-rot""s' mnial haittitity toward blaek eandidates's To

;.*pt this theory *outa in".6"t" the goals CongressjoJlght

t ,.ii"r" tiffidi"ting the lntent teJt of Mobilev. Bolil'en,

ia6 u. s. ffi (tggo), and-*ould prevent tnlnorlty voteis who

t"n".t""tty b""n denied an opportunity to eleef represent-

"iir"r 
of ilreir eholce from establishing a eritieal element of a

vote dilution elaim.
In amending $2, Congress rejected the requlrem-e1t 11-

,r*."d bt th"is Cburt {n Bokl'en, 8W, that !2 plaintlffs

**t p"ou" the diserlminatory lntentof st'ate or loeal govern-

,.ri" i, adopting or maintaining the ehallenged eleetoral

;il;t.r. "' Apiellants' suggestion that the discriminatory

irt".t 
"ii"aivtarial 

whlte vot-ers must be proVen ln_order to

rn"t 
" 

out a $2 claim must fail fot the very reasohs Congress

i"i".t"a tttiinhnt test with respect to governmehtal bodies.-Sl" 
nn*rt."m, The Reinearna[ioh of the Intent Standard:

il";fJrdg"s and At-Large Electlon Cases, 28 How' L' J'

4e6 (1e86).--1'fr" 
S"ii"te Report stated that one reason the Sehate Com-

hil;;;";doned the lntent test wes that "the Committee

:-. . h;";J persdasive testimony,that the intent test is unnee-

;;*r,; dfislve beeause lt inv6lves eharges of raeism on the

! lt ls true, as we heve reeognlzed prevloyslr, that raelal hoatillty mey

"rt"i 
ir"ii".iel btoc votlng. bnibd-tcwish Organizalian; t. Cowy, 130

ii.*ti.'il"i,-i6 tinzx n:&nt v. Loitse,.affi -!: 
s', at 623' But'. as we

iipii, iiir,lil;trio;, rhJecguat mo6-verlon of the vorer har no relevanee

a;;" dilutlon cletm. Thte ls not to suggest' that raelal btoe votlng ts

mee-neutrat; because voter behavlor *rtet"t"s wlth reee' obvlously

It ls not. lt should t" t,,r"'t"""a, though' ss one oommentator har ob

r"*J, U,ii ttlhe ebsenc" of ratt"t enlmus ls but one elemen[ of raee

rr"ototity." Note, Geometry and Geography 2O8'

'The Senate Report t"dJih"-"d''"ni th"t the worde "oh aceount

of r'ee.,, contslned ln '! qa), create any requlrement of purpoeeful

iir.i,nir"rirrl:irF it p"ffiv [elearl t'hai congrese has used.the words

ffiffiil;'o"" ri -for'lnihe Aet to mean'wlth reepect tolrale o1

*iri, iiJ rrt to .onnoitiny tcqulred purpoae of reelal dlscrlmlnatlon'"

S. ReP. Tl-8, n. 109.

THORNBURG o' GINGLES !!9

part of individual officials or entire eommunities'" S' Rep'

16- The Committee found the testimony of Dr' Arthur S'

fi"r.irg, Chairman of the United States Commission on

Civif nigfits partieularly persuasive, He testified:

"'[Under an intent testJ [llitigators representing ex-

.ruJ.d minorities will have to explore the motivations of

individual couneil members, mayors, and other citizens'

The question would be whether their decisions were

-oiirltua by invidious raeial eonsiderations' Such

lnquiries can only be tlivisive, threatening to destroy any

"*i.[irg 
raeial piogress in a eommunity. It is the intent

test, nit the resuits test, that woultl make it neeessary

ir-f."ra individuals as racist in order to obtain judieial

retief."'
IDdd. (foottrote omitted).--Til 

g"rr" threat to raciat progress and harmony whieh

C";;":;;"r."ir"d from requiring proof that racism eaused

tn"-ia"ptfln or mainten"n." of a challenged ejectoral mecha-

ni., i. p..sent to a much greater degre.e in the proposed re-

orrir"r"* that plaintiffs Jemonstrate that racial animosity

i"i.trir"a whil,e voting patterns' Under the old intent

t"rt, pf"f"tiffs might ,o."."Ld by provlng o-nly that a limited

nr.fl. of elected offfcials were raeist; under the new intent

i".f pf"irliifs would be required- to.prove that most of the

,f,it" .o.runity is racist in order to obtain judieial relief.

lf is difneutt to imagine a more racially divisive reqttirement'

Aseeondreason-Cong'"''rejeetedtheoldintenttestwas
ttrat in most eases it plaied an,iinordinately diffleult burden"

;i, plaintiffs. iua. The new intent test would be

"qr"tty,'if 
not more, burdenLgm"' - In order to Plove that a

iinriir' faclor-racial hostility-determined white voters'

ililrt* ii *orfd be neeessary to demonstrate that other

*t"ntirttv relevatrt cansal factors, sueh as soeioeeonomie

il;;;;l;;itties and candidate expenditures, do not eorrelate

["[i"" tfr,r" raeial animosity wittrwhite voting behavior' As

one eommentator has exPlained:



THORNBURG U. GINGLES

"ilIany of thetsel independent varlables . . . would be

all but impossible for e soeial seiehtist to operatlonalize

as InUervai-level independent variables fof use in a tnultl-
ple regresslon equatibn, rrhdther on a step-wise basls or
not. hti cotrduet sueh an ettensive statistleal analyals

as this implles, moreover' ean beeome prohlbltlvely
expensive.

;'Cotnpared to thle sbrt of effoitl pfoving dlserlmlna-
tory tntent ln the edoption of an atJarge system ts bgth

simple irnd inexpenslv0.', MeCrary, Diserlrnlnatoly In-

teni: The Contlnuing Relevanee of "Purpose" Evldenee

ln Vote-Dilutton Lawsuits, 28 How. L.l. 463, 4gz
(1986).

The ffnal and most dlapositlve reason the Senate Report

repudiated the old intetrt test was that lt "aske thi wrohg

qulstlon." S. Rep. 36. Amended $2 esks instead "whether
minoritles have equal aeoe88 to the proees8 of eleeting thelr
representatives." Ibid.

^tr"oeusing 
on the dlserlmlnatory lnteht of the voters, rather

than the behavior ofthe voters, aleo asks the wrong queation'

All that matt6rs under $2 and under a funetlonal theory of
vote dilution ls voter behavior, not its explanatlons' More-

over, aB we have explained in detail, suptt., requlring proqf

that raelal eorrsiderailons ietually cansed, votei behavlor will
r"iutt-.orrtrary to eongressionai lntent-ln situetions where

e btaet mlnorlt] that frrretionatly has been totally exeluded

froln tfr" politieal pnoces8 will be unable to establish a $2

vlotatlon. The senatc Report's remark eoneerning the old

lntent test thus ls pertlnent to the new teat: The requlrement

thata"eourt. . . make ddepurate. . . ftndlng of lntent, after

aeeeptlng the proof of the faetors lnvolved ln the whitn lv.
neaiater, ll2- U. S. ?661 enalysis . . . [wouldJ serlously

etirtal tire proepeets of eradlea*ng the remalnlng lnetenees

of raiat dlserimlnatlon ln Amerleen electlons." Id., ,,;37.
We therefore deeline to adopt sueh a requirement'

THORNBURG u. GINGLES

6

Summary

ln sum, we would hold that the legal eoneept of racially po-

larized voting, as it relates to elaims of vote dilution, refers
only to the existenee of a correlation between the raee of vot-
ersahd the seleetion of eertain eandidates. Plaintiffs need

hot firove eausation or intent in order to prove a primafacie
ease of raeial bloc voting and defendants may not rebut that
case with evidehee of eausation or intent.

IV
THE LECAL SIGNIFICANCE OF SOME BLACK

CANDIDATES' SUCCESS

A

North Cafolina and the United States maintain that the
Dietriet Court failed to aeeord the proper weight to the

8ueee88 of Some blaek candidates ln the challenged districts.
Blaek resldents of these distriets, they point out, aehieved

improved representation in the 1982 General Assembly elec-

tion.r They also note that blaeks in House District 23 have

eqioyed proportional representation consistently since 19?3

"ni 
-th"t 

bfacks in the other distriets have oeeasionally

eqfoyed nearly proportional representation.i This electoral

dt

.T'he retevant resutts of the 1982 Generat Aesembly election are as

followa. Houae Dlstrlct 21, ln whleh blecks make up 21.8% of the popula-

tlon, eteeted one btaek to the elx-person Houee delegatlon. Houee Dlstriet

28, in whleh blaekr eonetltute 86.8% ofthe populatlon, elected one blaek to

the three-persotr Houee detegatlon. In Houee Dlstriet 36, where blaeks

onrtltute 90.6% ol the population, one bleek was eleeted to the elght-

member delegatlon. ln House Dlstrlet 89, where 26.1% of the population

h btaek, two blaeka were eleeted to the f,ve'member delegation' In Se1-

ate Dh[rht 22, where blaeke c{nstltute 24.3% of the populatlon, no black

war elected to the Senate ln 1982.
.Ilre unlted stetee polnta out that, under a substantlally ldentieal

predeeearor to the ehallenged plan, see n. 16, rtpto, House Dlstrict 2_l

itected e btack to lte elx-member delegatlon ln 19&), House Dietrlct 39



42 THORNBURG u. GINGLES

Bueeess ciemonstrates eonelusively, appellants ahd the United
States argue, that blaeks in those distriets do not have "less
opportunity than other members of the eleetorate to partiei-
pate in the politleal process and to eleet tepresentatives
of thelr eholee." 42 U. S. C. $ 19?S(b). Essentially, appel-
lants and the United States eontend that lf a raelal minority
gains proportiohal or hearly proportional iepresentatlon lri e
slngle eleetion, that faet alone preeludes, as I matter of lawl
flnding a $2 violatlon.

Section 2 provides that "[tlhe extent to whleh members of
a proteeted elass have been elected to offiee . . . ls one eir-
eurnstahee whieh rnay be eohsidered." 421J. S. C. $ lg?S(b).
The Senate Committee Report also idetrtifles the eitent to
whieh minority eandldates have aueeeeded as a pertinent fae-
ior. S. Rep. 29. Hosever, the Senate Report expressly
states that "the election of a few minorlty eandidates does not
'neeessarily foreelose the possibility of dilution of the blaek
vote,"' nd,ting that if lt did, "the possibility exlsts that the
majorlty eltizens might evade t$21by manipulating the elec-
tion of a'Bafe'minority eandidate." Id., aL29, n. l16, quot-
ln! Zirnmer v. McKeithnn,486F. %J lzffi, f307 (CA6 1973)
(en bane), aff'd s?rb nam. Eaat Canoll Parish School Board,
v. Maishnll,4UU. S. 6:16 (lyi6) (per curiam). The Senate
Committee deeided, lnstead, to "'require an independent
eonsideration ofthe reeord."' S. Rep. 29, h. l16. The Sen-
ate Report also emphasizes that the question whether "the
politieal proeesses are'equally open'depends upon a seareh-
ing praetieal evaluatlon of the 'past and present reality."'
1d,., al30 (footnote omitted). Thus, the language of $2 and
its legislative history plainly demonstrate that proof that
some minority eandidates have been eleeted does not fore-
elose a $2 elaim.

Moreover, in condueting its "independent consideration of
the reeord" and its "searehing praetieal evaluation ofthe'past

elected e blaek to ltb hve-member delegetlon ln l9l4 and lg16, and Senete
Dlstrlet,zz had a blaek Senator between 1916 end 1980.

TH0RNBURG u. GINGLES 4s

and present reality,"' the Distriet Court eould appropriately
take aeeount of the eircumstanees surrounding recent black
eleetoral sueeess in deciding its signifieanee to appellees'
elaim. In partieular, as the Senate Report makes eletr, Id.,
.at 29, n. 116, the eourt could properly notiee the fact that
black eleetoral sueeess increased markedly in the lg82 elec-
tioh -an electioh that oceurred after the instant lawsuit had
been filed-and could properly eonsider to what extent "the
pendeney of this very litigation [might have] worked a one-
time advantage for blaek candidates in the form of unusual
organized political support by white leaders eoneerned to
forestall single-member districting."" 690 F. Supp., at 36?,
h.27.

Nothing in the statute or its legislative history prohibited
the eourt from viewing with some eaution blaek candidates'
sueeess in the 1982 eleetion, and from deciding on the basis of
all the relevant eireumstanees to aeeord greater weight to
blaeks' relative laek of sueeess over the eourse of several
recent eleetions. Conseqrrently, we hold that the District
Court did not err, as a matter of law, in refusing to treat the
faet that some blaek eandidates have sueeeeded as dispositive
of appellees' $ 2 claim. Where multimember distrieting gen-
erally works to dilute the minority vote, it cannot be de-
fended on the gtound that it sporadically and serendipitously
beneftts minority voters.

nSee also Zinmler v. Mcl{eith,e,n, 485 F. 2d 1297, 1307 (CAti 1973)
("[WJe eannot endorse the view that the suceess of black candidates at the
polls neeessarily foreclosea the possibillty of dilution of the black vote.
Sueh suecees might, on oecasion, be attrlbutable to the work of politicians,
who, apprehending that the support of a black eandidate would be politi-
eally expedient, eampaign to insure his election. Or such success might be
attributable to politieal aupport motlvated by rlifferent eonsiderations-
namely that election of a black candidate will thwart successful challenges
to electoral schemes on tlilution grounds. In either situation, a eandidate
eould be eleeted despite the relative politieal baekwartlrress of black
resldents ln the eleetoral distriet").



THORNBURG u. GINGLES

B

The Distriet. Court did err, however, in ignoring the

signlfleance of the su,stnined 8ueee88 blaek votert have ex-
peiieneed in House Distriet 28. In that distriet, the last six

Lleetions have resulted ln proportional representation for
blaek residents. This persistent proportional representation
is ineonsistent with appellees' allegation that the ability of
black voters in Distriet 2.3 to eleet representatives of their
ehoice is not equal to that eqioyed by the whlte majorlty'

In some situations, it may be possible for $ 2 plaintiffr to

demonstrate that sueh sustalned sueeess does not aecurately

refleet the rninority grdup's obility to eleet its preferred

representatives,il but appellees have not done so here. Ap-

o"-ll""r Dresented evldenee relating to blaek eleetoral suecess

in the last three eleetion; they failed utterly, though, to offgr
env exblanation for the sueeess of blaek eandidates in the
pr"rtotit three eleetions. Consequently, we believe that the

bi.trt.t Court erred, as a matter of lawl in lgnoring the

sustained Buceess black votere haVe enjoyed in House Dis-

trlet 23, end would reverse with respeet to that Distriet'

v
ULTIilATE bETEBMINATION OF VOTE DILUTION

Finally, appellants and the United States dispute the

Distriet couri,s ultimate eonelusloli that the tnultlmember

dl.t"i"tlng seheme at iesue ln this ease depriv"4 tJ"qk votefs

of an equil opportunity to partleipate ln t[re irclitleal proeess

and to eleet representatives of their ehoiee.

A

As an initial mattei, both North Carolina and the United

states eontend that the Distriet court's ultimate eonelusion

that the ehallenged multimember disttiets operabe to dilute

' r we have no ottsslon ln thls ease to dectde wha[ typee of lpeclel elreum-

stinees eoutd eatlefaetorlly demonstrete that eustdned euecese does not ae-

."iii"tv refleet the mlnority'a ablltty to eleet ltr prefeiicd repreaentatlves.

TTIORNBURG u. GINGLES 45

black citizens' votes is a mixed question of larv and fact
subject to de noao review on appeal. [n support of their
proposed standard of review, they rely primarily on Bose

Corporation v. Consumers Union of U. S.,Izc., 466 U. S.

485 (1984), a ease in which we reeonfirmed that, as a matter
of eonstitutional law, there must be independent appellate
review of evidence of "aetual maliee" in defamation eases.

Appellants and the United States argue that because a frnd-
ing of vote dilution under amended $ 2 requires the applica-
tion of a rule of law to a partieular set of facts it constitutes a
legal, rather than faetual, determination. Reply Brief for
Appellants ?; Brief for llnited States as Andcus Curiae
18-19. Neither appellants nor the United States cite our
several precedents in whieh we have treated the ultimate
finding of vote dilution as a question of fact subject to the
elearly-erroneous standard of Rule 62(a). See, e. 9., Rogers
v. Lodge, d68 U. S., at 622-627; City of Rome v. United,
States, d46 U. S. 166, iAs (tgSO); White v. Regester, 412

U. S., aL 765-470. Cf. Anderson v. Bessemer City, 470

U. S. _, _ (1985)

ln Regester, su,pra, we noted that the District Court had
based its eonelusion that minority voters in two multimember
tlistriets in Texas had less opportunity to participate in the
politieal proeess than mqiority voters on the totality of the
eireumstanees and stated that

"we are hot inclined to overturn these flndings' repre-
senting as they do a blend of history and an intensely
local appraisal of the design and impact of the . . .

multimember district in the light of past and present
reality, politieal and otherwise.'1 Id., aL769-770-

Quoting this passage from Regester with approval, we
expressly held in Rogers v. Lodge, supra, that the question
whether an atJarge election system was maintained for dis-
eriminatory irurposes and subsidiary issues, whieh include
whethef that system had the effect of diluting the minority
vote, were questions of fact, reviewable under Rule 52(a)'s



46 THORNBURG u. GINGLES

elearly-erroneous standard. 458 U. S., at 622-623. Simi-
larly, in City of Rome v. I|nited Slntes, we deelared that the
question whethbr eertaih eleetoral struetures had a "diserimi-
natory effeet," in the sense of diluting the minority vote, was
a question of faet subJeet to elearly-erroneous review. 446

U. S., at 188.

We reaffirm our view that the clearly-etroneous test of
Rule 62(a) is the appropriate standard for appellate review of
a flnding of vote dilutlon. As both amended $2 and its legis-
lative history make elear, in evaluatlng a statutory elaim of
vote dilution through distrieting, the trial eourt ie to eonsider
the "totality of the eireumstances" and to determine, based

"upgn a searching praetieal evaluation of the 'past and
preseht reality,"'S. Rep. 30 (footnote omltted), whether the
politieal proeess is equally open to minority voters. "'This
determination is peeuliarly dependent trpon the faets of eaeh

ease,"' Rogers, sryra, al62l, quoting Neaett v. Sidas,6?l F.
2d 2W, n4 $A619?8), and fequires "an intenaely loeal ap-
praisal of the design and impaet" of the eontested eleetoral
meehahisms. 458 U. S., at622. The fact that amended $2
and lts legislative history provide legal standards whleh a
eourt musl apply to the facts in order to determine whether

!2 has been violated does not alter the standard ofrevlew.
Aa we explained in Bose, Rule 62(a) "does not inhibit an 8p-

petlate eourt's power to eomect etrom of law, including those

that may infeet a sG.ealled mixed flnding of law and fact, or n
ffnding of faet that is predieated on h hisuhderstanding trf the
governing mle of law.)' 466 U. S., at 601, elting Pullmntt'
Stnndnd v. Swinl,466 U. S. 2?3, 28? (f982); Inwood' I'ab'
otatories,Itu. v. Irses Lhbotetories,Iru.,466 U. S. 84{, 866,

n. 16 (1982). Thus, thb applieatlon of the elearly-erroheous

s[andard to ultimate ffndlngs of vote dilution preserves the

beneftt trf the triel eodft's pdftieular familiariti wlth the

lndiSenous politleal teality withriut ehdangerlng the rule of
law.

TIIORNBURG u. GINGLES

B

The District Court in this ease earefully considered the
totality of the cireumstanees and found that in each district
raeially polarized voting; the legaey of offfeial diserimination
ln voting matters, education, housing, employment, and
health serviees; and the persistenee of campaign appeals to
raelal prejudice aeted in eoneert with the multimember dis-
tricting seheme to impair the ability of geographically insular
and politieally eohesive groups of black voters to participate
equally in the political proeess and to eleet eandidates of their
ehoiee. It found that the suceess a few black candidates
have enjoyed in these distriets is too reeent, too limited, and,
with regard to the 1982 eleetions, perhaps too aberrational,
to disprove its eonelusion. Exeepting House District 23,
with respect to whieh the District Court committed legal
error, aee suprt,, at 

-, 
we afflrm the Distriet Court's judg-

ment. We eannot say that the District Court, composed of
loeal Judges who are well-acquainted with the political reali-
ties of the State, elearly erred in eoneluding that use of a
multimember eleetoral strueture has eaused black voters in
the dlstriets other than House District 23 to have less oppor-
tunity than white voters to elect representatives of their
ehoiee.

The Judgment of the District Court is

Alfirmed in part, reuersed, in parl.



APPENDIX A

Pereentages of Votes Cast by Blaek and Whlte Voters fof
nlaek Carrdidates in the. FIve Contested Dlsttiets

Sawte Diatrict ee

Primnrg Getoml
Wt*tn Blank Whiln Blonk

THORNBURG U. GINGLES

47 87
8?8g2 &l

1978 (Alexander)
l98o (Alexander)
1982 (Polk)

l9I8 (Blue)
1980 (Blue)
1982 (Blue)

i978 Serui,te
Barns (Repub.)

19f8 HuBe
Clement
Spaulding

1980 Hune
Spaulditrg

1982 Hune
Cletrietrt
Spaulding

21 76 nla
Bl 8l 44

39 82 46

41 94
nla nla
:t3 9.1

nle
90
9l

Hmne District et

Primary Gennal
Whiti Btaek lfl*ii;te Blotk

Houe Dishiil*S
Primnry Gennml

whitn Btack White Black

nla

hla

l7

l0 89
16 s2

, zli 32

3? 90

nla nla
37 89

904en/e

nla n/d
43 89

THORNBURG o. GTNGLES

House District 36

Primnry
White Black

lg80 (Maxwell) 22 7l
1982 (Berry) 60 ?9

1982 (Rlchardson) 39 7l

House District 39

Primnry
White Btack

1978 Houe
Kennedy, H.
Norman
Ross
Sumter (Repub)

1980 House
Kennedy, A.
Norman
1980 Sennle
Small
1982 House
Hauser
Kennedy, A.

690 F. Supp., at 369-371.

49

General
Wlr,ite Black
?8 92
42 92
29 88

General
White Btack

93
nla
nla
25

96
nla;

32
n/e
nla
33

32
nla

76
29
53
nla

86
36

6l

28
8

t7
nla

40
18

12

42 87
46 94

26 80
36 87



m . THoRNBURG,. GTNGLbS

APPENDIX B

iilaek Candidates Eleeted iiront ? Originalty Contested

Distriet
(No.Sebts)

House 8 (4)
House 21 (6)

House 23 (3)

House 36 (8)
Hotrse 39 (6)

Senate 2(2)
Senate n G,

Dlstrietg

Frlor 0o

ts7i, lyll ls74 10?6 lvrS 1e80

See Brief for Appellees, table printed between pages 8 and 9;

App. es-e4.

000000
000001
0lll11
000000
001100
000000
00tll0

te82
0
I
I
I
2
0
0

SUPREME COURT OF TIIE TJNITED STATES

No. 83-1968

LACY H. THORNBURG, ET AL., APPELLANTS U.

RALPH GINGLES ET AL.

oN AFPEAL Fnou Tup UNITED STATES DISTRIC"T CoURT FoR
THE EASTERN DISTRICT OF NOBTH CAROLINA

lJune &), 19861

Jusrtcr WIIITE, eoneurring.
I join Parts I, II, III-A, III-8, IV-A and V of the Court's

opinion and agree with Justrca BnnNNltt's opinion as to
Part IV-B. I disagree with Part III-C of Jusrtcn Bnru-
NAN'B opinion.

Jusrtcn BRruNlN states in Part III-C that the crucial
factor in identifying polarized voting is the race ofthe voter
and that the race of the candidate is irrelevant. Under this
test, there is polarized voting if the majority of white voters
vote for different eandidates than the m{ority of the blacks,
regardless of the raee of the candidates. I do not agree.
Suppose an 8-member multimember district that is 60Vo

white and 40Vo blaek, the blaeks being geographieally loeated
so that 2 safe blaek eingle-member districts could be drawn.
Suppose further that there are 6 white and 2 black Demo-
erats running against 6 white and 2 black Republicans.
Under Justtcn BngNNAH's teat, there would be polarized
votlng and a likely $ 2 violation if all the Republicans, inelud-
lng the 2 blaeks, are eleeted, and 80% ofthe blacks in the pre-
dominantly blaek areas vote Democtatic. I take it that there
would also be a violation in a single-member district that is
60% black, but enough of the blaeks vote with the whites to
eleet a blaek eandidate who is not the ehoice of the mqjority of
blaek voters. This is interest-group polities rather than a
rule hedging against raeial disoimination. I doubt that this



THORNBURG u. GINGLES

is what Congress had in mind in amending 02 as it did, and it
seems quite at odds with the diseussion in Whibortb v.
Chnois, 403 U.S 124, 149-160 (1971). Furthermore, on
the faets of this ease, there ls no need to draw the
voter/eandidate dlstinctlon. The Distrlet Court dtd hot and
reaehed the corfeet result exeept, itr tny view, with rgspeet to
Distriet zg. .

ST]PREIIIE COURT OF THE TJNITED STATES

No. 83-1968

LACY H. THORNBURG, ET AL., APPELLANTS u.

RALPH GINGLES ET AL.

ott AFpnAL FRoM THE uNrrED srATEs Drsrnlcr couRT roR
tHn restnRN DISTRICT oF NoRTH cARoLINA

[iune 80, 1986J

Justtce O'CoNNoR, with whom THs CHrsr Jusrrcr, Jus-
TtcE POwELL, and Justtco RnHNqutst join, eoncurring in
the judgment.

In this ease, we are called upon to eonstrue $ 2 of the Vot-
ing Rights Aet of 1965, as amendedJune29,1982. Amended
$ 2 is intended to eodify the "results" test employecl in
Whitcomb v. Clnuis, 403 U. S. 124 (1971), and White v.
Regester,4l2U. S. 765 (19?3), and to reject the "intent" test
propounded in the plurality opinion in Mobile v. Bolden, 446
U. S. 66 (i980). S. Rep. No. 9?-41?,pp.27-28 (1982) (here-
lnafter S. Rep.). Whereas Bold,en required members of a
raeial minority who alleged impairment of their voting
stiength to prove that the ehallenged electoral system was
ereated or maintained with a discriminatory purpose and led
to dlseriminatory results, under the results test, "plaintiffs
may ehoose to establish discriminatory results without prov-
ing any kind of diseriminatory purpose." S. Rep. 28. At
the same time, however, $ 2 unequivoeally disclaims the
ereatlon of a right to proportional representation. This
diselaimer was essential to the eompromise that resulted in
passage of the amendment. See id., aL 193-194 (Additional
Views of Sen. Dote).

In eonstruing this eompromise legislation, we must make
every effort to be faithful to the balance Congress struek,
This is not an easy task. We know that Congress intended



2 THORNBURGU. GINGLES

to allow vote dilution eiaims to be brought under $ 2, but we
also know that Congress did not intend to ereate a right to
proportional representation for minority voters. There is an
inherent tension between what Congress wlshed to do and
what it wished to avold, beeause any theory of vote dilutlon
must neeessarily rely to Bome ettehf oii a measure of minor-
ity voting strength that makes some felerence to the propor-
tion between the minority group and the eleetorate et large.
ln additlon, several important aspects of the "results" test
had reeeived little attention in this Court's caseg or ln the de-
eislons of the Courts of Appeals employing that test on whleh
Cohgress also relied. See id., at 32. Speelffeally, the iegal
tneaning to be given to the coneepts of "raeial bloe votlng"
and "minority voting atrength" had been left largely trrt-
addressed by the eourts when $2 was amended.

The Court attempts to tesolve all these difffeulties today.
Flrst, the Court supplles deflnitions of raeial bloe voting and
minority voting strength that wlll apparehtly be applieable in
all eases and that will dietate the struetufe of vote dilution
litlgation. Seeond, the Court adopts e test, based on the
level of minority eleetoral Bueoess, for determining when an
electoral seheme hae sulfieiently diminished minorlty toting
strength to eotrstitute vote dilutlorr. Thlrd, although the
Coutt does not aeknowledge lt etpressly, the eomblnatlort of
the Court's deflnitlon of minority voting strehgth and its test
for vote dilutlotr results ln the ereatlon of a rlght to a foim of
proportlonal representatlon ln favor of all geogrdphleally and
politieally eohesive tnlnority groups that are lafge enough to
eohstitute rnaJoritles lf eoncentrated within one or more sin-
gle-member dlstriets. In so doing, the Cotlrt hae dlsre-
garded the balanee etrueh by Congress ln amending $2 and
has failed to apply the results test ss desetibed by thls Court
ln Whitcomb and Whitn. ,, 

I

ln order to explain my disagreement with the Coutt's ln-
terpretation of $ 2, it ls useful to illttstrate the impact that el-

TTIORNBURG u. GINGLES B

ternative districting plans or types of districts typically have
on the likelihood that a minority group will be able to elect
eandidates it prefers, and then to set out the critical elements
of a vote clilution elaim as they emerge in the Court's opinion.

Consider a town of 1,000 voters that is governed by a coun-
cil of four representatives, in which 307o of the voters are
blaek, and in whieh the black voters are eoncentrated in one
section of the city and tend to vote as a bloc. It would be
possible to draw four single-member districts, in one of whieh
blaeks would constitute an overwhelming majority. The
blaek voters in this district would be assured of electing a
lepresentative of their choiee, while any remaining black vot-
ers in the other districts would be submerged in large white
inajorities. Thls option would give the minority group
roughly proportional representatlon.

Alternatively, it would usually be possible to draw four
single-member distriets ln two of which black voters eon-
stituted mueh harrower majorities of about 607o. The black
voterr in these distriets would often be able to elect the
representative of their choice in each of these two districts,
but if even 20Vo of the blaek voter's supported the eanditlate
favored by the white mlnority in those tlistriets the eandi-
dates preferred by the rnajority of black voters might lose.
This option woutd, depending on the eireumstances of a par-
tieular eleetion, sometimes give the minority group more
than proportional representation, but would inerease the risk
that the group woultl not aehieve even roughly proportional
representatlon.

It would also usually be possible to draw four single-mem-
ber distriets in each of which blaek voters eonstituted a mi-
nority. In the extreme ease, blael< voters woulcl eonstitute
30% of the voters in eaeh distriet. Unless approximately
30% of the white voters in this extreme ease baeked the mi-
nority candidate, blaek voters in such a district would be
trnable to elect the eandidate of their ehoiee in an. eleetion
between only two eandidates even if they unanimously sup-



THORNBURG o. GINGLES

ported him. This option would make it difffeult for blaek
voters to eleet eandidates of their ehoiee even with signiffeant
white support, and all but impossible without sueh support.

Finally, it would be possible to eleet all four represent-
atives in a single atJarge eleetion in whieh each voter eould
vote for four eandidates. , Under this scheme, white voters
eould elect all the representatives even ifblack voters turned
out in large numbers and voted for one and only one eandi-
date. To illustrate, if only four white eandidates ran, and
eaeh received approxintately equal support from white vot-
ers, eaeh would reeeive about ?00 votes, whereas blaek vot-
ers eould east no more than 300 votes for any orre eandidate.
If, on the other hand, eight white eandidates ran, and white
votes were distributed less evenly, so that the ftve least fa-
vored white eandidates reeeived fewei than 300 votes while
three others reeeived 400 or more, it would be feasible for
blaeks to eleet one representative with 300 votes even with-
out substantial white support. If even 25% of the whlte vot-
ers baeked a partieular minority eandidate, and blaek voters
voted only for that eairdidate, the candidate would reeeive a
to'tal of 476 votes, whieh would ensirre victory unless white
voters also eoneehtrated their votes on four of the elght
remaining eandidates, so that eaeh received the support of
almost 70% of white voters. As these variations show, the
atJarge or multimember distriet has an inherent tendeney to
submerge the votes of the mlnority. The minority group's
pfospects for electoral sueeess under sueh a distrlet heavily
depend on a variety of faetors sueh as votet turno[t, horf
many eandidates run, how evenly white support is spread,
how mueh white support is given to a eandidate or eandidates
preferred by the minority group, and the extent to whieh mi-
ndrity voterr engage in "bullet voting" (which oceurs when
voteri refrain from easting all their votes to avoid the rlsk
that by voting for their lowe|-ranked ehoiees they riray give
those eandida0es enotigh toteu to defeat their hlgher-ranked
ehoiees, aee o,nle, 

"6 -, 
tr. 6).

THORNBUITG u. GINGLES 5

There is no difference in principle between the var.ying
effeets of the alternatives outlined above and the varying
effects of alternative single-district plans and rnultirnember
districts. The type of distrieting selected and the way in
whieh distriet lines are tlrawn ean have a powerful effect on
the likelihood that members of a geographically and politi-
eally eohesive minority group will be able to eleet candidates
of their choiee.

Although $2 does not speak in terms of "vote dilution," I
agtee with the Court that proof of vote dilution can establish
a violation of $ 2 as amended. The phrase "vote dilution," in
the legal sense, simply refers to the impermissible discr.imi-
natory effect that a multimember or other clistricting plan
has when it operates "to eancel out or minimize the voting
strength of raeial groups." White, 412 U. S., at 761-r. See
alao Fortson v. Dorsey, 379 U. S. 433, 439 (1965). This defi-
nition, however, eoneeals some very formidable rlifficulties.
Is the "voting strength" of a racial group to be assessed solely
with referenee to its prospeets for eleetoral suceess, or should
eourts look at other avenues ofpolitical influence open to the
racial group? Insofar as minority voting strength is as-
sessed with reference to electoral sueeess, how should undi-
luted minority voting strength be measured? How muclr of
an impairment of minority voting strength is neeessary to
prove a violation of $ 2? What eonstitutes racial bloc voting
ahd how is it proved? What weight is to be given to evi-
denee of aetual electoral sueeess by minority candirlates in
the face of evidence of raeial bloe voting?

The Court resolves the first question summarily: minority
voting strength is to be assessed solely in terms of the rninor-
ity group's ability to eleet eandidates it prefers. Ante, aL

Under this approaeh, the essenee of a vote di-
lution elaim is that the State has ereated single-member or
tnultlmember distriets that unaeeeptably impair the minority
group's ability to eleet the candidates its members prefer.



6 THORNBURG {,. GINGLbS

ln order to evatuate a elaim that a partieular multiriretnbet
distriet or single-member distrlet has diluted the mlnority
group's voting strength to a degree that violates $ 2, how-
evet, it is also heeessary to eotrstruet e measure of "irndi-
tuted" minority Votlng strength. "[T]he phrase [vote dilu-
tlon] itself suggests a horm tvith fespect to whieh the faet of
dilution may be aseertalned." Mississippi Repu,blican Erec-
uthte Cotttrniltne v. Brooks, 469 U. S. 1002, l0l2 (1984)
(RgHHeutsr, J., dissenting from Bummary afffrmanee). Put
simply, ln order,to deelde whether an eleetoral system has
tnade it haider for minorlty Voters to eleet the eandldates
they prefer, d eourt thrlst have sn ldea in mind of how hatd it
"should" be for minority toteft to eleet thelr preferred eandi-
dltes uttdet dn aeeeptable sysbern.

Several possible rieasutes of "undiluied" minorlty voting
strength suggeot themselves. Flrst, a eourt eould simply
rlse piopottionality ds lts guide: if the minority group eonsti-
tuted SOga of the voters in a givert area, the eourt tvould re-
gurd the minorlty group as having the potentlal to elect 80%
of the representatives in that area. Second, a eourt eould
posit sotne alternatlve distrieting plan es a "normal" or "falr"
eleetoral seheme ahd attempt to ealeulate how many eandl-
dates preferred by the minority group would ptobably be
eleeted under that schetrte. There are, as We have Beeh, e
variety of ways ltr whlch even single-member dlstrlcts eould
be drawn, and eaeh will pteaent the minorlty group wtth lts
own array of eleetoral rlsks and beneflts; the eourt inight,
thetefore, cotrsider d fhnge ofaeeeptable plens in atternptlng
to estimate "undiiuted" mlrioriti voting strength by thls
mbthod. fhtrd, the eourt eould attempt to ai'rive at d plan
that wouid maximiie feasible minority eleetoral sueeess, and
use this degree of predlcted Bueeess as lts tneasure of "undi-
tuted" minorlty votlni strength. lf a court were to efnploy
this third alternative, it would often faee hard ehoiees about

what would truly "mexlmlze" tninofity eleetoial Bueeess. Atr
example is the seenario described above, in whleh a mlnority

tHORNBURGu. GINGLES ?

group eould be eoneentrated in one completely safe district or
divided among two districts in eaeh of whieh its members
would constitute a somewhat preearious majority.

The Court today has adopted a variant of the third ap-
proaeh, to wit, undiluted minority voting strength means the
maximum feasible minority voting strength. In explaining
the eleinents of a vote dilution elaim, the Court ffrst states
that "the ininority group must be able to demonstrate that it
is sufHciently large and geographieally eompact to eonstitute
a majority in a single-member district." Ante, al 

-. 
If

not, apparently the rninority group has no eognizable claim
that ite ability to elect the representatives of its choiee has
beetr impaired.r Seeond, "the minority group must be abte
to show that it ts politieally eohesive, that is, that a signift-
cant proportion of the minority group supports the same ean-
didates." Ante, at 

-. 
Third, the Court requires the mi-

rl exprees no vlew as to whether the ebillty of a minnrity group to eon-
etltute a mqlority ln a alngle-member dletrlet should constitute a threehold
requlrmen[ for a elalm that the uee of multimember districts impairs the
ablllty of mlnorlty voters to partieipate ln the politieal processes and to
eleet repiesentatives of their cholee. Because the plaintiffs in this case
rfould meet that requirement, iilndeed it exlste, I need not decide whether
It lc lmpoeed by 0 2. I note, however, the artllleiality of the Court's dis-
tlnetlon between elalms that a mlnority group's "ability lo elect the repre-
sentatlveg of fltsl eholee" has been lmpaired and elaims that "ite ability to
itfiucrce eleetlons" hes been impalred. Ante, aL 

-, 
n. 12. lt is true

that a mlnorlty group that could eonstitute a mqlority in a single-member
dletrlet ordlnarlly has the potential ability to elect representatives without
whlte aupport, end that a mlnority that eould not eonstitute such a rnqjority
ordlnarlly does not. But the Court reeogrrizee that when the eandidates
preferred by a mlnorlty group are elected in a multimember district, the
mlnority group has el,ected Lhose eandidates, even if white support was in-
dlspensable to these vlctoriee. On the same reasoning, if a minority group
that ls not large enough to eonstitute a votlng maJority in a single-member
dlstriet ean show that white eupport would probably be forthcoming in
some sueh dietrlet to an extent that would enable the election ofthe candi-
detes its members prefer, that minority group would appear to have dem-
onatrated that, at least uirder thie meesure oflts votlng strength, it rvould
be able to eleet aome candidates of lts cholee.



THORNBURG o. GINGLES

nority group to "demonslrah that the white majority votes
sufffeiently as a bloe to enable it-in the absenee of speeial
eireumstances . .,. 

-usually to defeat the mitrority's pre-
ferred candidate." Ibid. If these three requirements are
met, "the minority group demonstrates that submergence in
a white multimember distriet impedes its ability to eleet its
ehosen iepresentatlves." Ibid. That is to say, the minority
group has proved vote dilution in violatlon of ! 2.

The Court's deflnitiotr of the elements of a vote dilution
' elaim is siniple and invariable: a eourt should ealeulate minor-
ity voting strength by uisuming that the minority group is
eoneentrated in a single-member distrlet in whlch lt eonsti-
tutes a votlng inqfority. Where the minority group is not
large enough, geographieally coneehtfated enough, ot poiitl-
eally eohesive enough fot this to be possible, the minority
group's elaim fails. Where the tninority gro[p mbets thesd
feqlrirements, the represehtativee that it eould eleet ln the
hypothetieal distriet or districts in whieh lt eonstitutes a
tnaJority will serve as the measure of its undiluted votlng
strength. Whatever plan the State aetually adopts fnust be
assessed in terms of the effeet it has on this undiluted votlng
strength. Ifthis is indeed the single, universal shandard for
evaluating undiluted minority voting strength for vote dilu-
tion purposes, the standard is applieable whether what ls
ehallenged is a multimember district or a partieular single-
member distrleting sehenie.

The Court's statement of the elemehts of a vote dilutioh
elaim also supplies ah answer to another question posed
above: hmt ruunh of ari impairment of undiluted minority vot-
ing strength is neeessary to prove vote dilution. The Court
requires the minority grdup that satisfies the threshold re-
quirements of size and eohesiveness to prove that lt will
usually be unable to eleet as many representatives of lts
ehoiee under the ehallenged distrieting seheme as its un-
diluted voting strength would permit. This requirement,
then, eonstitutes the true test of vote dilution. Agaln, no

THORNBURG u. GINGLES I

reason appears why this test would not be applicable to a
vote dilution claim challenging single-member as well as
multimember distriets.

This measure of vote dilution, taken in eoqiunction with
the Court's standard for measuring undiluted minority voting
strength, creates what amounts to a right to ltsunl, roughly
proportional representation on the part ofsizeable, eompact,
eohesive minority groups. If, under a particular multi-
member or single-member distriet plan, qualified minority
groups usually earrnot eleet the representatives they would
be likely to elect under the most favorable single-member
distrieting plan, then $ 2 is violated. Unless minority
sueeess under the ehallenged eleetoral system regularly ap-
proxlmates this rough version of proportional representation,
that system dilutes minority voting strength and violates $ 2.

To appreeiate the implieations of this approaeh, it is useful
to return to the illrtstration of a town with four eouneil repre-
sentatives given above. Under the Court's approach, if the
blaek voters who eomprise 30Vo of the town's voting popula-
tion do not usually sueeeed in eleeting one representative of
their ehoiee, then regardless of whether the town employs at-
large eleetions or is divided into four single-member districts,
its eleetoral system violates $ 2. Moreover, if the town had a
black voting population of 407o, on the Court's reasoning the
blaek minority, so long as it was geographically and politi-
eally eohesive, would be entltled usually to elect two of the
four representatives, sinee it would normally be possible to
efeate two districts in which blaek voters constituted safe
majorities of approxim*ely 807o.

To be sure, the Court also requires that plaintiffs prove
that racial bloe voting by the white majority interaets with
the challenged distrieting plan so as usually to defeat the mi-
nority's preferred eandidate. In fact, however, this require-
ment adds little that is not already eontained in the Court's
tequirements that the minority group be politically eohesive
and that ifs preferrecl eandidates usually lose. As the Court



l0 THORNBURG u. GINGLES

aekrtowledges, under its approaeh, "in general, a white bloe
vote that normally will defeat the eombined etrength of mi-
nority support plus white'erossover'votes rises to the level
of legally signiffcant white bloc voting." Ante, at 

-. 
But

this is to deffne legally signifleant bloe voting by the raeial
tnqiority in terms of the, extent of the raeial minority's elee-
toiul sueeess. If the minority can prove that lt eould consti-
tute a mqjorlty in a aingle-membei district, that it strpported
eertaln eandidates and that those eandidates have not usually
been eleeted, then a ftnding that there is "legally signlfleant
white bloe voting" will neeessarily follow. Otherwise, by
definition, those eandidates would usually have won rather
than lont.- 

A;il;p"d by the Court todey, thetr, the basle eontouts of a
vote dilution elaim reqdire no referenee to tnost of the "Zim-
m,er faelors" that were developed by the Fifth Cireuit to
implement White's results test and whieh were highlighted
in the Senate Report. S. Rep. 28-29; see Zimmer v.

McKeith,en, 486 F. 2d 1297 (1973), aff'd stcb nnm. East
Canoll Pwish School Board, v. Marshnll, 424 U. S. 636
(19?6) (per anriam). If a minority group ls politieally and
geographieally eoheslve and large enough to eonstitute a vot-
lng mqiolity in ohe or mofe single-ineinber districts, then un-
less whtte voters usually suppoit the rninority's pieferfed
eandidates in sufneient numbers to ehable the minotity group
to eleet as many of those eandidates as it could elect in such

hypothetieal distrlets, tt will routinely follow that a vote dilu-
tlon elaim ean be made out, and the multimember distriet rfill
be invalidated. There ls slmply no need for plaintiffs to es-

tablish "the history of votlng-related diseriminatlon in the
State or politieal subdivlslon ," Q,nln, s[ 

-t 
dr "the extent to

whieh the State or politieal subdlvision has used voting prae-

tiees or procedures that tehd to ehhanee the opportunity for
diserlmihation agalnst the minority group," dbdd., or "the ei-
elusion of metnbera of the mlnority group fronl eatrdidate slat-
ing pioeessea," ibid,., or "the extent to whieh minotity group

THORNBURG u. GINGLES ll
members bear the effeets of past discrimination in areas sueh
as education, employment, and health," ibid., ot'"the use of
overt or subtle raeial appeals in political campaigns," ibid., or.
that "eleeted offieials are unresponsive to the particularized
needs of the members of the minority group. " Ante, at 

-.Of eourse, these other factors may be supportive of such a
elaim, beeause they may strengthen a eourt's confidence that
minority voters will be unable to overeome the relative disad-
vantage at which they are plaeed by a particular districting
plan, or suggest a more general laek of opportunity to partici-
pate in the politieal proeess. But the faet remains that elec-
toral gueeess has trow emerged, under the Court's standard,
as the linehpin of vote dilution elaims, and that the elements
of a vote dilution elaim ereate an entitlement to roughly pro-
portional iepresentation within the framework of single-
member districts.

II
In my view, the Court's test for measuring minority voting

strength and its test for vote dilution, operating in tandem,
eome closer to an absolute requirement of proportional repre-
sental,ion than Cong'ess intended when it codified the results
test ln $2. It ls not neeessary or appropriate to decide in
this case whether $ 2 requires a uniform measure of undiluted
minority voting strength in every ease, nor have appellants
ehallenged the standard employed by the District Court for
assessing undiluted minority voting strength.

In this eese, the Distriet Court seems to have taken an ap-
proaeh quite similar to the Court's in making its preliminary
assessment of undiluted minority voting strength:

"At the time of the creation of these multi-memb"r di.-
triets, there were eoneentfations of blaek citizens within
the boundarles of eaeh that were sulffeient in numbers
and eohtiguity to constitute effective voting majorities in
slngle-member distriets lying wholly within the bound-
aries of the inulti-rnember districts, which single-mem-
bet distriets would satisfy all eonstitutional require-



12 THORNBURG o. GTNGLES

ments of poputation, and geographieal eonftguration."
Gingles v. Efuni8lfttt, 690 F. Supp. 346, 368-369 (ED
N. c. le84)

The Court goes well beyond strhply sustaining the Distrlet
Couit's deeisiori to dmploy this measure of undiluted minority
vbting strength ds h reasonable one that is tonslstent with
$2. In my view, we should refrairt from deciding in this ease

whether a court must inverlably posit as its measure of "un-
. diluted" minority voting atrength single-membef distriets in
which minority group members eonstltute a majority.
Thete is substantial doubt that Congress intended "undilul,ed
minofity voting strength" to mean "maximum feasible minor-
ity voting strength." Even if that is the appropriate deffni-
tion in some eireumstanees, there is no lndieation that Con-
gress intended to mandate a single, universally applicable
standard for measurlng undiluted tninority voting strength,
regardless of local eonditions and regardless of the extent, of
past diseriminatlon against minority votert in a partieular
State or poiitieal suMivisioh. Sinee appellants have not
raised the issue, I would assume that what the Dietriet Court
did here was permissible under $ 2, and leave open the
broader question whether 92 requires this approaeh.

What appellants do eontest is the propriety of the District
Cotrt's stahdard for fote dilution. Appellants elalm that the
District Court held that "[a]lthough blaeks had aehleved eon-

siderable Bueeess lrt rtinning state legislatlve seats ln the
ehallenged disttieis, thelr failufe to eonsistently attaln the
number of seats lhnt numberc alone wr,ld' presum,ptiaely
gbe th,em(d. e., ln proportion to their presenee irt the popula-

tioh)," standing alone, constituted a violatlon of $ 2. Brief
for Appellants 2O (elnphasls in original). This holdlng,. ap-
pellents afgue, elearly cohttavenes $ 2's provlso that "nothlng
ih this seetioil estsblishei a flght to heve nieinbeta of i pro'
tected elass eleeted'ln numbers equal to thelr proportion in
the populatioh.'j 42 U. S. C. $ i9?3.

rlr

THORNBURG o. GINGLES lB

I believe appellants'eharaeterization of the District Court's
holding is ineorreet. In my view, the District Court eon-
cluded that there was a severe diminution in the prospects
for black eleetoral sueeess in each ofthe challenged distriets,
as eompared to single-member districts in whieh blacks could
eonstitute a mqiority, and that this severe diminution was in
laige part attributable to the interaetion of the multimember
form of the distriet with persistent racial bloc voting on the
part of the white majorities in those districts. See 690 F.
Supp., at372.2 The Distriet Court attaehed greahweight to
this eireumstanee as one part of its ultimate finding that "the
ereation of edeh of the multi-member districts ehallenged in
this aetion results in the black registered voters of that dis-
triet being submerged as a voting minority in the district and
thereby having tess opportunity than do other members of
the eleetorate to participate in the political proeess and to
eleet representatives of their choice." 1d,., aL374. But the
District Court's extensive opinion clearly relies as well on a
variety of the oLher Zimnter faeLor-s, as the Court's thorough
summary of the Distriet Court's ftndings indicates. See
antn, aL

If the District Court had held that the challenged multi-
member distriets violated $ 2 solely beeause blacks had not
eonslstently attained seats in proportion to their presenee in
the population, its holding would clearly have been inconsist-
eht with $ 2's diselaimer of a right to proportional representa-
tion. Surely Cdngtess did not intend to say, on the one
hatrd, that metnbers of a protected class have no right to pro-
portlonal representation, and on the other, that any eonsist-
ent failure to aehieve proportional representation, without

rAt tlmes, the Distrlct Court aeema Uo heve looked to simple proportion-
atlty tather than [o hypothetieal slngle-member distrleto in whieh blaek
votefg woirld eonstltute d mqlority. See, d. g., 690 F. Supp., Ct 36?. No-
where ln lts oplnlon, however, did the Dlstrlct Court state that g 2 requires
that thlnorlty groups eonslstently attain the level of eleetoral euccese that
would eorrecpond wlth thelr proportlon of the total or voting population.



I
I

I
'l

l4 ' ' THoRNBU[c o. ctNct Es

more, violates $ 2. A requirement that tninority representa-
tion usually be propoitional to the minority group's propor-
tioh in the populatlon is not quite the Balrle as e fight to strict
proportional representation, but it eomes so elose to sueh a
right as to be lneonsistent with $2'e diselainier and with the
results test that is eodifled ln ! 2. In the words of Senator
Dole, the archlteet of the eompromise that resulted iri pas-

sage of the amendments to $2:

"The language of the s[bseetloh explicltly rejeets, as dtd
Whito and its progeny, the hotioh that members of a pro-
teeted elass hive a tight to be elecied itt numbers equal
to their proportiorl of the populatlon. The extent to
whieh members of e proteeted elass have been elected
under the ehallenged practlce or strueture isJust one fae-

toi, amohg the totality of eireumstanees to be eonsid-

ered, and-is not dispositive.'i S. Rep. lg4 (Additiorral
Views of Sen. Dole).

On the Bame reaaonlng, I would teJeet the Court's test fot
Vote dilution. The Court measures undiluted mlnority vot-

. lng strehgth by referenee to the possibility of creating single-
member aistri.ts in whieh the minority group would consti-

tute e tn{ority, rather than by iooking to.faw proportlonality
alone. The Court's standard for vote dilution, when eom-

bined with its test for undiluted minority voting strength,
makes aetionable every deviation from usual, tough propor-
tionatity in representatlon for any eohesive minority group.as

to whieh this degtee of ptoportionality is feasible within the
framework of Jingle-mem-ber distriets. Requirlng that
every minority group that eould possibly eonstitute a tn1l9t;
ity in a single-member distriet be assigned to sueh a dlstriet
would approaeh a requirement of proportional representation
as n"arly as is possible within the framework of single-mem-

ber distrlets. Sin." the Court's analysis entitles every sueh

minority group usually to eleet as many representatlves

uhdei f,multimember dietriet as lt eould eleet under the most

i"rri"Uf" slngle-tnetnber distrlct seheme, lt follows thet the

THORNBURG u. GINCLES l5

Court is requiring a form of proport,ional representation.
This approach is inconsistent with the results test and with
$ 2's diselaimer of a right to proportional representation.

In enacting $2, Congress eodifted the "results" test this
Court had employed, as an interpretation of the Fourteenth
AmendmenL, in White and Whitcomb. The factors devel-
oped by the Fifth Circuit and relied on by the Senate Report
simply fill in the eontours of the "results" test as described in
those deeisions, and do not purport to redeflne or alter the
ultimate showing of diseriminatory effeet required by Whit-
cornD and ilVhite. In my view, therefore, it is to Whitcomb
and White that we should look in the ffrst instance in deter-
mining how great an impairment of minority voting strength
ls required to establish vote dilution in violation of $ 2.

The "results" test as reflected in Whitcomb and White re-
Quires ah lnquiry into the extent of the minority gtoup's
opportrinities to participate in the politieal proeesses. See
White, suprfl, at 766. While electoral Bueeess is a central
part of the vote dilution inquiry, White held that to prove
vote dilution, "it is not enough that the raeial group allegedly
discrlminated against has not had legislative seats in propor-
tion to its voting potential," 412 ll. S., at ?66-?66, and
Whitcomb flatly rejeeted the proposition that "any group
with distinetive interests must be represented in legislative
halls lf it is numerous enough to command at least one seat
and represents a minority living in an area sufficiently eom-
paet to eonstitute a single member district." 403 U. S., at
166. To the contrary, the results test as deseribed inWhite
fequires plaintiffs to establish "that the politieal proeesses

leading to nomination and eleetion were not equally open to
participation by the group in question-that its members had

,less opportunity than did other residents in the tlistrict to
participate in the politieal proeesses and to elect legislators of
their ehoiee." 412U. S., at 766. By showing both "a his-
tory ofdisproportionate results" and "strong indicia oflack of
politieal power arrd the denial of fair representation," the



16 THORNBURG u. GINGLES

plaintiffs ln White met this standard, which, as emphasized
just this Term, requires "a substantially greater showing of
adverse effeets than a mere lack of proportional representa-
tion to support a ffnding of uneonstitutional vote dilution."
Dauis v. Bandnm.err 

- 
U. S. 

-, -, - 
(i986) (plrt-

rality opinion).
Whetr Cohgress ainehded $ 2 it lntended to adopt this "re-

sults" test, while abandoning the edditiortal showlng of dis-
eriminatory intent required by Bold.en. 'i'he vote dilution
analysis adopted by the Court, tgday elearly bears little
resemblence to the "results" test that emerged in Whitcomb
and While. The Court's test fot vote dilution, eoinbined
with its standaid for evaluating "voting potential," While,
Bupro, at ?66, means that any raeial mitrority with distinctive
interests must us?rollg "be represented in legislative halls lf
it is numerous ehoitgh to eommand at least one seat and rep-
resents a mlnority living in ah area suffieiently eompaet to
eonstitute" a voting mqiority itt "a single member distriet."
Whitcomb,40s U. S.r at 166. Nothing lnWhitcomb,White,
oi the language ahd legislative history of $2 suppotts the
Oourt's creation,of this right to usual, roughly proportlonal
representatlon on the part of every geographically eompaetl
politieally eoheslve minorlty group that is large enough to
form a maJority in one or more sitrgle-tnember distrlets.

i would edhere to the approaeh outlined ln Whitcomb and
White aird followed, {,ith some elaboiation, ln Zimm,er and
other eases in the Courts of Appeals prior to Bolden. Under
that approaeh, a eourt should eonsidei all relevant faetors
bearing on whether the minority group has "less opportunity
than other members of the eleetorate to partieipate in the po-

litleal proeees d,nd, to elect repfesentatlves of thelr eholee."
42 U. S. C. $ 1frI3 (embhasis added). The court should hot
foeus solely otr the mlnority groupts ability to elect represent-
atives of its choiee. Whatever measure of undiluted rhinor-
Ity voting stfertgth the eouft employs in eonneetion wtth eval-

uatlrtg th1 nres;trce or absenee of minoflty eleetoral 8ueee88,

ri.l

THORNIIURG u. GINGLES l7

it should also bear in rnintl that "the power to influence the
political proeess is not limited to wintring eleetions." Daais
v. Bandetner, sltpra) at 

-. 
Of eoutse, the relative laek of

minority eleetoral sueeess under a challenged plan, when
eompared with the sueeess that would be predicted under the
measure of undiluted minority voting sttength the court is
employing, can constitute powerful evidenee of vote dilution.
Moreover, the tnitrority group may in faet lack aeeess to or
influence upon representatives it did not support as candi-
dates. Cf. Daais v. Band,emer, su,pra, x[ 

- 
(Pownll, J.,

eohcurring in part and dissenting in part). Nonetheless, a
reviewing eourt should be required to {ind more than simply
that the minority group does not usually attain an undiluted
lneasure ofeleetoral suecess. The court must ffnd that even
substantial minority sueeess will be highly infrequent under
the ehallenged plan before it may eonelucle, on this basis
alone, that the plan operates "to eaneel out or minimize the
voting strength of [theJ racial grou[p]." White,412 U. S.,
at ?65.

III
Only a plurality of the Court joins Part III-C of Jusrtcn

BRrNNAl.t's opinion, which addregses the validity of the sta-
tistieal evidenee on which the Distriet Court relied in finding
raeially polarized voting in each of the ehallenged districts.
Insofar as statistieal evidence of divergent racial voting pat-
terns ls admitted solely to establish that the minority g"otlp
is politieally eohesive and to assess its prospects for eleetoral
Bueeess, I agree with the plurality that defendants eannot re-
but this showing by offering evidenee that the divergent ra-
cial voting patterns may be explained in part by causes other
than taee, sueh as an underlying divergence in the interests
of minority and white voters. I rlo not agree, however, that
such evidehee ean hever affect the overall vote dilution in-
qirlfy. Evidenee that a candidate preferred by the minority
group in a partieular eleetion was rejeeted by white voters
ior reasons other than those which made that candidate the



18 THORNBURG U. GINGLES

preferred ehoice of the minority group would seem elearly
relevant in answering the question whether bloe voting by
white voiets will eonsistently defeat minority eandidates.
Sueh evldenee would suggest that anothef eatrdidate; bqually
preferred by the mlnorlty Broop, tnight be able to uttraet
greater white support ln future eleetions. '

I believe Congress also intended that explanatlonc of the
reasons why white voters rejected minority eandidatee would
be probative of thd likelihood that eandidates eleeted without
deeisive minority support would be willing to take the mlnor-
ity's interests into aeeount. In a eommunity that is polatized
along raeial lines, raeial hostility may bar these ahd other in-
dlreet avenues of politieal lnfluenee to a mueh greatef extent
than in a community where raeial animoaity ls abseht al-
though the interests of raeial groups diverge. Indeed, the
Senate Report elearly stated that one faetor that eould have
probative value in $ 2 eases was "whether there ls a slgnifl-
eant laek of responsiveness on the part of eleeted offfcials to
the p'articularized needs of the members of the minority
group." S. Rep. 29. The overall vote dilution inquiry nel-
ther requires nor permits an arbitrary rule against eonsider-
ation of all evidenee eoneeining voting preferenees other than
statistical evideneb of raeial voting patterns. Sueh a nrle
would give no effeet whatever to the Senate Report's re-
peated emphasis on "intensive faeial polities," on "raeial
political eonsiderations," and on whether "raeial politles . . .

dominate the eleetoral pioeess" as one aspeet of the "racial
bloe voting" that Cohgress deemed ielevant to showing e $ 2

violation. S. Rep. 83-34. Similarly, I agree with Jusrtcp
Wumr that the plurality's eonelusion that the raee of the ean-

didete ls always irrelevant ln identifyihg racially polarized
voting eonflicts wlh Whitcomb and is hot neeessary to the
dlspositiotr of this ease. Ante, ht 

-' 
(WnItn, J., eoneur-

fing in part and ltr the judgrnent and dissenting ln pert).
ln this,ease, Eri the Cotrrt grudgngly acknowledges, the

Dlsttiet Cotirt etearly erred in aggreoating deta ftom all of

THORNBURG u. GINGLES

the ehallenged distriets, and then relying on the faet that on
average, 8l.7Vo of white voters did not vote for any black
eandidate in the primary elections selected for study. Ante,

"g -, 
n. 28. Although Senate District 22 eneompasses

iiouse District 36, with that exception the distriets at issue in
this ease are distributed throughout the State of North Caro-
lina. White ealls for "an intensely loeal appraisal of the de-
sign and impact of the . . . multimember distrlet," 412 U. S.,
at 769-770, and racial voting statietics from one district are
ordinarily irrelevant in assessing the totality of the cireum-
stanees in another distriet. In view of the speeific evidenee
from each district that the District Court also considered,
however, I eannot say that its conelusion that there was se-

vere raeial bloe voting was clearly emoneous with regard to
any of the challenged distriets. Except in House District 23,

where raeial bloe voting did not prevent sustained and virtu-
ally proportional minority electoral sueeess, I would aecord-
ingly leave undisturbed the District Court's decision to give
great weight to racial bloe voting in each of the challenged
districts.

IV

Ilaving made usual, roughly proportional suecess the sole
focus of its vote dilution analysis, the Court goes on to hold
that proof that an oeeasional minority eandidate has been
eleeted does not foreelose a $ 2 claim. But JusttcE BREN-
NAN, Joined by Jusrtcr Wuttr, concludes that "persistent
proportional representation" will foreelose a $ 2 claim unless

the plaintiffs prove that this "sustained sueeess does not ae-

eurately refleet the minority group's ability to elect its pre-
ferred representatives." Ante, aL 

-. 
I agr:ee with Jus-

TICE BRENNAN that eonsistent and sustained success by
candidates preferred by minority voters is presumptively in-
eohslstent with the existence of a $ 2 violation. Moreover, I
dgree that this case presents no oecasion for determining
what would eonstitute proof that sueh suecess did not aecu-



20 THORNBURGU. GINGLES

rately refleet the minority group'B aetual voting strength in a
ehallenged distriet or distriets.

In my view, the District Court erred in assessing the ex-
Uent of blaek electoral Buseess in House Distriet 89 and Sen-
ate Distriet 22, aa well as in House District 23, where the
Court aeknowledges error. As the evidenee summarized by
the Court in table form shows, ante,6[, 

-, 
App. B, the de-

gtee of blaek elbeboral sueeess dilfered widely ln the seven
originally eontested districts. In House Dlstrict 8 and Sen-
ate Distrlet 2, neither of whieh is eontested ln thls Court, no
blaek eandidate had ever been elected to the offiees ln ques-
tion. In llouse Dlstrlct 21 and Houee Distriet 36, the only
lnstahees of blaek electoral Bueeegs eatne in the two tnost re-
eent electiohs, ohe of whleh took plaee during the pendenef of
thls litigation. By eontrast, in House Distrlct 39 and Sehate
Distrief 22, blaeli Buccrioses, although lntermittent, dated
baek to 19?4, and a blaek candidate had been elected in each
of these distriets in three of the lest flve eleetions. Fitrally,
ln House Dlstrlet 23 a bleek eandidate had beeh eleeted ih
eaeh of the last six eleetions.

The District Court, drawing no distinetions atnong these
distriets for purposes of lts flndings, eoheluded that "[tlhe
overall resulta aehleved to date at ell levels bfelective offlce
rire tninimai in relatloh to the pereentage of blaeks in the total
populatlon.' 690 F. Supp., at 36?. The Distriet Corlrt
elearly erred to the eitent that lt eorrsldered eleetoral sueeess

in the aggregate, rather than in each of the ehallehged dis-
triets, sinee, as the Court states, "[t]he inquiry into the exiet-
eiree of vote dilution is distriet-speeifle." Ante, g[ 

-tn. 28. The Court assertg that the District Court was free to
regard the results of the 1982 eleetions with Busplclon ahd to
deeide "oh the basis of all the relevant eiretrmstartees to ae-

eord greater weight to blacks' relatlve laek of sueeess ovef
the eourse of ieveral reeent etectionsl" ante, at 

-, 
but the

.Court does not explain how thls tbehtrique worlld apply irr

Senate Distriet 22, whete a blaek eandidate was eleeted irt

THORNBURG u. GINGLES

three eonseeutive eleetions from 1974 to 1978, but no black
eandidate was elected in 1982, or in House District 39, where
blaek candidates were eleeted in lg?4 and 1976 as well as in
1982. Contrary to what the District Court thought, see 690
F. Supp., at 367, these pre-1982 suecesses, which were pro-
portional or nearly proportional to blaek population in these
three multimember distriets, eertainly lend som,e support for
e flnding that black voters in these distriets eqjoy an equal
opportunity to participate in the political proeess and to elect
representatives of their choiee.

Despite this error, I agree with the Court's eonelusion
that, except in House District 23, minority electoral suceess
was not suflieiently frequent to eompel a flnding of equal
opportunity to participate and elect. The District Court
found that "in eaeh of the challenged districts raeial polariza-
tlon in voting presently exists to a substantial or severe de-
gtee, and . . . in each distriet it presently operates to mini-
mize the voting strength of blaek voters." 1d,., at 372. I
eannot say that this ftnding was clearly erroneous with re-
speet to House Distriet 39 or Senate District 22, particularly
when taken together with the District Court's findings eon-
eerning the othet Zitnmer faetors, and hence that court's ul-
timate eonelusion of vote dilutlon in these districts is atle-
qtrately stipported.

This flnding, however, is clearly er?oneous with respect to
House Distrlet 23. Blaeks eomprise 36.3?o of the population
in that distriet, and eonstitute?.8.8% of the registered voters.
ln each of the six eleetions sinee 1970 one ofthe three repre-
sentatives frotn this distriet has been a blaek. There is no

ftnding, or any reason even to suspeet, that the sueeessful

blaek eandidates in Distriet 23 did not in faet represent the
interests of blaek voters, and the District Court did not ftnd
that blaek Buceess in previous eleetions was aberrant.

Zimm,et'a eaveat against necessaiily foreclosing a vote di-
lutloh elaitn on the basis of isolated black successes, 486

F. 2d, at 130?; see S. Rep. 29, n. ll5, eannot be pressed this



n TITORNBURG u. GINGLbS

far. Indeed, the 23 Court of Appeals deeisions on whieh the
Senate Report relied, and whieh are the best evldence of the
seope of this eaveat, eontain no example of minority eleetoral
Buecess that even-temotely approximates the eonsistent, dee-
adeJong pattern in Diatrlet 21. See, e. 9., Tu.ra,q v.

MeKeithen 490 F. 2d 191 (CAE 1973) (no blaek eandidates
eleeted); Wallace v. Hur,ae,616 F. 2d 619 (CA6 1975), va-
eated on other grounds, 426 U. S. 947 (19?6) (one blaek eandi-
date elected).

I do not propose that eonsistent and virtually proportional
minority electoral Bueeess should always, as a matter of law,
bar flnding a $ 2 violation. But, as a general rule, sueh sue-
eese is entitled to great weight in evaluating whether a ehal-
lenged eleetoral meehanism has, on the totality of the eireum-
Btanees, operated to deny blaek voters an equal opportunity
to partieipate in the politleal proeess and to eleet represent-
atives of their ehoiee. With respect to House Distriet 23, the
District Court's failure to aceord black electoral sueeess sueh
weight was clearly erroneous, and the Distriet Court identl-
fled no reason for not giving thls degree of suecess preelusive
effeet. Aeeordingly, I agree with JusricE BRENNAiI that
appellees failed to establish a violation of $ 2 in Dlstriet 2.3.

v
When members of a reeial minority ehallenge e multl-

member distriet on the grounds that it dilutes their votlng
strength, I agree wlth the Court that they must show that,
they possess sueh atrength and that the multimember distriet
lmpairs it. A eorrt must therefore appraise the mlnority
group's undiluted voting strength in order to assess the
effeets of the multimember district. I would reserve the
question of the proper method or methods for making thls as-

sessment. But onee sueh an assessment is made, in my view
the eyaluation of,an,alleged impairment of voting strength
requireg,eonsideration of the minority group's aeeess to the
politieal prccesses generally, not solely eonsideration of the

THORNBURG u. GINGLES 23

ehanees that its preferred eandidates will actually be elected.
Proof t,hat white voters withhold their support from minor-
ity-preferred candidates to an extent that consistently en-
sures their defeat ie entitled to signillcant weight in plaintiffs'
favor. However, if plaintiffs direet their proof solely to-
wards the minority group's prospects for eleetoral sueeess,
they must show that substantial minority sueeess will be
highly infrequent under the ehallenged plan in order to estab-
lish that the plan operates to "eaneel out or minimize" their
votlng strength. White, 412 U. S., at 765.

Compromise ls essential to much if not most major federal
legislation, and eonfidenee that the federal eourts will enforee
sueh eompromlses is indispensable to their ereation. I be-
lieve that the Court today strikes a different balance than
Congress intended to when it eodifled the results test and
diselaimed any right to proportional representation under $ 2.
For that reason, I join the Court's judgment but not its
oplnlon.



STJPREME COURT OF THE I]NITED STATES

No. 8!l-1968

LACY H. THORNBURG, ET AL., APPELLANTS u.

RALPH GINGLES ET AL.

ON APPEAL FROM THE UNITED STATES DISTRICT COURT FOR
THE EASTERN DISTRICT OT NORTH CAROLINA

lJune 30, 19861

Jusrtcr StovrNs, with whom Justtcn MeRstrlll and
Jusrrcn BlAcruun Join, coneuming in part and dissenting in
part.

In my opinion, the findings of the bistrict Court, which the
Court, fairly summarlzes, ante, at,4-8; 19-20 and n. 23;26-27
and nn. 28 and 29, adequately support the District Court's

Judgment eoneerning House District 23 as well as the balance
of that Judgment.

I, of eourse, agree that the election of one blaek eandidate
in eaeh eleetion sinee 1972 provides signifieant support for
the State's posltion. The notion that this evidenee ereates
some sort of a eonelusive, legal presumption, ante, * 43-44,
la hot, however, supported by the language ofthe statute or
by its legislative history.' I therefore cannot agree with the
Court's view that the District Court committed error by fail-
lng to apply a rule of law that emerges today without statu-

rsee ozte, at 42 ("Sectioh 2 provides that'ttlhe extent to whlch mem-
bers of a proteetEd class have been eleeted to olllce . . . is one eircumetance
whleh tnay be consldeted ,' 42 U. S. C. ! lg73(b). . . . However, the Sen-

ate Report expresely statee thet 'the eleetlon of a few minority eandidates
does not "neeessarlly foreelose the possibility ofditution ofthe blnck vote,"
notlng that lf lt dld, 'the poaslbility exlsts thet the mqiorlty eitizens mlght
evade [0 2] by manlpulatlng the election of a "Bafe" minority candi-
date.' . . . Ilre Senate Committee deeided, lnstead, io 'require an inde-
pendent eonslderetloh of the reeord"') (lnternal eitations omitted).



2 THORNBURGrl. GINGLES

tory support. The evidenee of eandidate sueeegs in Distriet
23 is merely one part of an extremely large reeord which the
District Court earefully considered before maHng lts ulfi-
mate findings of faet, all of whleh should be upheld uhder d
normal applieatlon of the "elearly efroneous" standard that
the Court traditlonally applies.t

The Court identifles fhe re"ron why the Bueeess of one
blaek eandidate ln the eteetlons In lgil8, 1g80, and lg82 is not
ineonsisteht with the Distriet Court's ultlmate flnding eon-
eerning House Dlstriet 23.t The fdet that one blaek eandi-
tiate was also eleeted til the lwz, lg7A, and 19?6 eleefiohs,
ante, App. B, ls not suffleient, in rny opinioh, to overeotne the
additional flndihgs that apply to Housl Distrlet %, Nwell as
to other districts in the State for eaeh of those years. The
Court aeeurately summarizes those flndings:

"The Dlstrlet Court ln this ease earefully eonsidered the
Cotality of thb elreurrtstanees and found that ln eeeh dis-
triet raeially polarized Voting; the legaey of offteial dis-
etiminatlon in voting matters, edueation, housing, em-
floyment, and heelth servlees; and the persistehce of
eampaign appeals to taelal preJudiee aeted ln cohcert
wlth the multimdmber dlstrlcting seheme Uo lmpair the
ability of geographieally insular ahd politleally eoheslve
groups of blaek votere to partieipate equally ln the politl-
eal proeess and to eleet eandidates..of.their eholee. it
found that the stieeess n few biach eandldates have 0h-
Joyed ln these distrlets ls too reeent, too limlted, arrd,
with regard to the 1982 eleetions, perhaps too aberra-
tional, to disprove lts eonelusion." Ante, at 46-47.

To paraphrase the Court's eonelusion about the other dis-
tti.t- ,b4 r I eannot say that the Distriet Court, eornposed of

2See anle, at a6 ftT]he applieatlot ofthe elearly-er?oneouo standard to
ultlma[e flndlngs of voUe dllutlon preserveo the beneftt of the trlel eotirt's
partleular femlllarlty wlth the'lndlgenoiie polltleal irality dthotrt endari-
gerlng the rule of lew").

'*e anil,c, at t9-20 and n. &l; Tl, n.29;42-41.

THORNBURGU. GINGLES a

toeal judges who are well-acquainted with the political reati-
ties of the State, elearly erred in eoneluding that use of a
rnultimember eleetoral strueture has eaused black voters in
House District 23 to have less opportunity than white voters
to elect representatives of their ehoiee.r Accordingly, I eon_
cur ln the Court's opinion exeept Part IV-B and ex-eept inso-
far as it explains why it reverses the judgment respecting
House Distriet 23.

i Even under the Court's analysis, the deelslon simpty to reverse-with-
out a femand-ls myetifylng. It lg also extremely unfalr. F.imt, the
Court does not glve reapondents an opportunlty to address the new legal
standard that the court llnds deeisive. second, the court troes not eien
bother to explaln the eontoure of that etandard, and why it was not eatis-
lled ln thls ease. cf. antn, at 44, n. 88 ("we have no oeeasion in this case to
decide what types ofspeclal eireumslanees eourd eatisfactorily demonstrate
that 6uetalned Buceess doea not accurately refleet the minority's ability to
eleet lts preferred repreeentatives"). Flnally, though eouched as 

" 
.on-.tu-

elon about a "matter of law," id., aL 44, the Court,s abrupt entry of judg_
ment for petltlonei! on Dletrlct 28 re{leets an unwilringne.. to gir" ilre nfi-
ttlet Court the respeet lt le due, partleularly when, as in thls ease, the
Dlatrlet court hae a demonstrated knowtedge antl expertlse of the entire
context that Congrese dlreeted !t to coruider.