Motion for Leave to File Brief Amicus Curiae and Brief Amicus Curiae

    Un^.,",

l,u

Lii





* lu/,put>
u-*/^/tel

,1





.r/b,*4
A-*1", a/ /4r^,A'

fr^)y





.N

/,*,a" ;

"JZ,-l tu;





ffiepre$@mtatqcn
an& [tednstrEetEstg
EssuJ0s

Edited by
Bernard Grofman
University of California, lrvine

Arend Lliphart
University of California,

San Diego

Robert B. McKay
Aspen Institute for Humanistic

Studies and Institute of
Judicial Adrninistration

Howard A. Scarrow
State University of New York

at Stony Brook

Afterword by
Helnz Eulau

LexlngtonBooks
D.C. Heath and CompanY .
Lexington, Massachusetts
Toronto



ae lssues

ointeC out the

'nsider, fcr ex-
,. ith trvo iotes,
al vole shares

f's concept of
,1, 'A) when a

s necessarY for
)ower at all (in
,r two members
nply that when

igned in such a

:er of votes he

rs in his district.
mty, New York

rntl', New Ycrk

,rnty, New York

e mathematical
district i\itcnls
:nportant omis-
1g at the district
i multirnember

an "oP:imal"
lt the clear iden-

rderstanding oP-

lat preserve in-
! MacRae, 1962:

:t of the number
: additional can-
Jdon ma)' act as

al!y jeopardizing
particular, they

aning for reelec-

Representatives,
ihe earlY' 70s, a

to run no fewer

:ptions, this Pro-

inois House rvas

ie legislature.

c ' .n. ".) The Consequences of., c,-:_._iO. eid. i .. -,r,..iii\,. j lJultimemberL ';;.r' Districting

Malcoim E. Jewell

As we enter the period of post-1980 reapportic:.::.:.-.:s. one of the oues:i.rns

that remains uncertain is whether the tiei',c ronard single-member
distriaing in state legislatures and city councils *ill continue, or uhether
some state legislatures will retain multimember c.,:i:ts and whether some

city councils will continue to be elected a:.::;:. One causr.'oi this
uncertainty, of course, is the confusing and :::::.::crory decisions that
come from the courts. At the same time, noi e:ala:r is known about the

consequences of various types of districting. The :l'::rs hnvq \a/1g51led u'ith
the question of whether multimember i:i::.::, :;scriminare .tj:.unst
minorities, particularly racial minorities; the'.' :.:'.: also been concerned
about the discriminatory intent of such ciisi;.::'. :-:. li rhe courrs rurn to
political science literature for evidence aboul ::.:.-::.:equences of r:rrirrus
districting patterns, they lvill find relatively Ii::.: :'..:ence, most of uhich
deals uith the narrow questicn of the impa;i ;: :.:::lcting on nunrerical
representation of racial and partisan minori:::s. I 

'-_"_:est 
that $e nc'c'd to

broaden the scope of our inquiry into the co::;:.-::.:3s of districring.
There are more than trvo possible districr.:.;::..::ns. At one ('\rreme

are single-member districts; at the other is ::.: ::-.arge election of all
members of a city council or all state legislators ::. . ::-.::opolitan countl'. In
between these extremes are several patterns: ::.: :-.:bination of single-
member and atJarge districts in a city councill s:."'l =ulrimember districts
(two or three members) in a legislature; and a:-.-;: :::1'council elcctic'rns
with members either residing in districts or ele;::: .:. ::imaries by districts.
The efferts of multimen:ber districting are a.:: :.::::enr depending on
whether or not each candidate runs for a spe;;:1.'r..'s: or piace.

The consequences ol districting patterns iji --:i -i.','the same uhctlter
the unit of analysis is a city council or a s'.;:i .;a;..iiure, although the
effects on minorities may be more severe in a.-.:.. :-.uncil. At thc state
legislative levcl there is ahvays a possibilit)'i]:":.:::::san or racill grtrup
that is a minority in one county may'be a ni:'::. .-.::te other. lnrrtltcr
diffcrence is that state legislaturcs often ct'rn::.:.: ':-:.:- and multinrcnrtrcr
districts, while city councils may combine s::_i.:-:.:nbc.r and ar-large
districts. lvly remarks are rclatcd spccificalll' :-' ::::-. .:gislatures, trccausc

129



130 Representation and Redistricting lssues

my research has beert focusecl olt them, but I believe most of them are perti-
nent to city councils as well. (A year ago I conducted22O interviervs with
legislators in nine states on various aspects of district representation.) I will
look at the consequenccs of districting for elections, campaign techniques,
communications with constituents, the service role of legislators, and
policy.

Elections and Campaigns

The studies cited by Grofman (chapter 12, this volume) that are the most
comprehensive and rigorous confirm that racial minorities are much more
likely to be proportionately representecl in city councils when single-member
districts are used. A review of southern state legislatures (Jewell, 1980)

shows that those lower houses having a substantial minority of black
legislators are ones that use single-member districts exclusively (or like
Georgia use them in the counties with the greatest black political strength):
Alabama, Gcorgia, Louisiana, South Carolina, Tennessee, and Texas. The
states with very few black representatives are all ones using large

multimember districts in metropolitan areas: Arkansas, Florida, Missis-
sippi, Virginia, and North Carolina. Moreover, in such states as Texas,

Tennessee, Georgia, and Louisiana, there is evidence of a sharp increase in
black representation (and of Mexican-American representation in Texas)

following the shift to single-member districts.
Several factors affect the impact of districting on minority representa-

tion. Racial minorities nrust be large enough and concentrated enough to
benefit from single-member districts-and what is large enough depends of
course on the number of districts in a city or county. Obviously what counts
is not the minority's proportion of the population, but its proportion of the
voting population, and this depends on the level of potitical organization of
the minority. In partisan elcctiotts, thc crtrcial factor may be the strength of
the minority bloc in the party primary (usually the Democratic one). We

also need to consider the impact that part) organizations and leaders may
have on primaries and elections. In cities or counties with at-large elections
and with strong organizations, leaders of minority groups have sometimes

been able to win places on a rvinning slate for their group's candidates. This
means, of course, that minority candidates can be elected only if they are

acceptable to the slate makers. Leaders of minority groups can sometimes

elect their candidates in at-large eleclions (wirhout places or positions) if
they can mobilize their members to engage in single-shot voting, but this
tactic sacrificc's their influence over the selection of the other candidates.

Studies of electoral effects hiive been devoted more to racial than to par'

tisan minorities, but in partisan ciry council and legislative races the effects



Consequences of Districting 131

of ntultimember districts on parlisan minorities may be important' One

significant consequence of single-member districting in southern state

[iirr"irr., has been to elcct more Republicans in the higher-income sec-

tions of metropoliran ""*,i.r. 
iiis probable that rhe development of single-

member distiicts has accelerated the grorvth of southern metropolitan

Republicanparties"oo.ni,un..dtheirabilitytorecruitcandidates.Theim.
pact of districting on the partisan balance in a legislature depends on the

strength of partisan ""ii"g 
patterns. In Indiana, for example, there is an

unusually srrong ouairiJn' of straight-ticket voting. when all state

L-;irl;" *.r" 
"l".t.d 

at large in N,rario, county (Indianapolis), it was nor-

nral for trre Marion c;;;t jeregation to shift from being all-Democratic to

all-Republican. Now itu,'.orrntv is divided into five three-member House

districts,andstraight-ticketvotingprevailsineachdistrict'InNorth
Carolina, however, ii.r. it less itraight-ticket voting; in the atJarge

legislative races in **"p"fii"n counties it is common to have members of

both parties elected from a county'

The tactics una .ori' of srate legislative campaigns are quite different in

single-member and ; i;s" multimember. distric-t:' l:,^:t:t"-"ember
districts most legislariu. .uiaiaures run their own campaigns. They raise

ir,.i, o*n funds ind builct small organizations of volunteers. They run per-

sonal campaigrrr, going doo'-to-aoit, frequenting the shopping centers and

other gathering ptaces] and speaking to neighborhood groups' There is a

maximum of p...onui.ontutt, and a minimum of emphasis on issues' Can-

didates may seek til;;;;;f organized groups and mav get some help in

thegeneralelectionfromthepartyorganization'buttheirsuccessdoesnot
depend heavily on such gto'p'' Except in the larger' more populous

districts (such as those in Callfo"ia, New York' or Ohio)' the cost of such

campaigns is relativelY low'--'--it 
Jpu,,ern is diiierent in large multimember and at-large districts, par-

ticularly in metropolitan areas. in such districts it is impractical for can-

didates to campaign door-to-door or to concentrate on neighborhood

gatherings urra *".ti.rgs. Insread it is^necessary to speak at group meetings

around the district unJutt.na rallies if they can be organized. It is also more

necessary to use media advertising, though television may.be prohibitively

expensive. In such Jirtrr.t. assistance of the party organization becomes

moreimportant,""agroupendorsernentsasstlmehigherpriority.
Newspaper .ndorr.rn.nt'1nuy also be significant' Candidates of the same

party often 
""rrrpuign 

u' u t"u*' and soinetimes a slate of candidates will

work together in a primary' The costs of campaigns are almost inevitably

higher in a multime,iiu.i ait,ri.t (contrclling for other variables)' although a

candidate who runs as part of a team may be able to limit those costs'

Which type of district is more conclucive to the reelection of in-

cumbents?Itisgenerallyassumedthatnamerecognitionismoreimportant



Representation and Redistricting lssues

in large multimcntber clistricts, rvherc the vc)tcrs llar,c (roublc kccping track
of nlany candidatcs and the incunrbcnl, norrnally has a heaclstart ip nanre
recognition. To win rcclection, horvever, thc incumbc-nt may have to retain
the support of thc parry organizaticl, and,/or inrerest gr-oups. ln a single-
member district rhe incurnbent probably hes a borcr oppoirunity ro buita
and maintain a political base and to nraintain contacts with and provide scr_
vices for constituenls, thcrcby bcconring c.iifiisult to beat. Legislators
generally sL'eln to bclieve that their prospects for rcelcction are better in
single-member districts.

Communicating with and Serving
Constituents

The relationships betrveen legislators and constituents are fundamentally
different in single-member and in large multimember districts. In the single-
member district, there is more direct personal contact with constituents and
less reliance on the media and on organizecl groups. Legislators who
originally campaigned door-to-door sometimes go back to ttri precincts to
talk with vorers in betrveen elections. Most of them find ptenty of oppor-
tunities to talk to neighborhood groups and ro attend loci runction, lirenif thel' are not invited to spcak) to mninrain visibility, ancl acccssibiliry. A
few legislators (as in California and Texas) rnaintain disrrict offices thar are
regularly staffcd.

Although the job of maintaining conracrs is generalry easier in singre-
member districts, there are some{imes probie:ns tliat trccur when district
lines do i:ot fc!lou' natural cor.rni'.r:iity. t:c,u:til;,r ics. Some lcg:isiativc r.listricts
in metropolitan counties do not incorporate existing communities and do
not have nlany organized groups; sonie d;srricts include bits and pieces of
different counties; some in rural areas are so vast geographically as to make
personal contact difficult. To the extent that boundary lines cut across the
communities that forrn the basis for org:unizltions and cornmunications
patterns (such as nenspapers), single-nrember ciistricls may creare more dif-
ficulties for representation than do murtimember districts.

In a multimember district, particularly a countywide metropolitan one,it is impractical for a legislator to try to maintain close personal contacts
with constituerlts or attend neighborhoocl meetings scaitered across the
district. The advantage of representing a lar-eer district, particularly one en-
compassing a whole county, is that many groups are organized at the city or
county level. In other words, the legislator can reach constituents through
groups, or at least to reach those rvho are interested enough in public affairs
to attend group meetings. If several legislators represc-nt i countywide
districr., they may try to divide up the rvorkload ancl the invitations ro speak

132



Consequences of Districti ng

at nleetings. Those t'ho are the most aggressive and effective speakers,

however, may get the most opportunities to develop visibility and contacts

through organized groups.

In rural palts of a state, most legislators find it easy to get space in

ntwspapers to publicize their viervpoints atld activities because there are

usually only one or two legislators representing any county. In a large

metropoilitan area, however, all legislators must compete for attention in

the press. It is the legislators rvho understand public relations bcst and those

rvho mt:st often bsl'ome involr.ed in controversial issues wlto are most likely

to get media attention. In metropolitan areas the problem of getting media

access is essentially the same for legislators rvhether they represent single- or

multimember districts. One difference is those representing multimember

districts are more dependent on the media because they lack the oppor-

tunities for personal contacts. Some single-membcr district legislators also

find that they can titilize local or suburban (usually rveekly) nc\\ spapers that

have circulation areas roughly coinciding with or including their district' In

a number of states, Iegislators from single-member districts are using

newsletters to reach constituents. Some of these are quite professionally

done, and some include questionnaires on issttes. Thcse are mailcd to consti-

tuents in states rvhere the legislature has authorized funds for postage, and

are distributed door-to-door where funds are lacking. tsut this method of
communication is usecl almost entirely in single-mentber districts.

Legislators in large multi-districts consider it not to be feasible.

A significant trend in recent years has been the growing service role of
legislators. As legistators become more visible and their staff resources in-

crease, constituetlts are increasitrgly turning to them for help in dealing with

state agtncies or even for help rvith fede ral or locitl ofiicials. Thesc demands

from constituents occur where legislators are visible, and particularly where

legislators have advertised their rvillingness to perform services and have

made themselves and their staffs available. While thcse demands are

greater, and are different in character, in lower-inconte districts, they are

present ln other types of districts as well-if the legislator encourages them.

But the constituency-service role is important only for those legislators who

represcmt single-member district.s.
It is clear that legislators representing large rnultimember districts

seldonr get requests from constituents for ihdividual assistance. They lack

the visibility and Sxrsonal contact that are necessary, and for the most part

they do not believe that this is a lcgitirnate part of their job. In fact, some of
them bclieve that one advantage of multimentber districting is that
legislalors do nol get servicc reqtlests. More research is needcd on city coun-

cils to find out how or wlrether conslituents get help in dealing with local of-
ficials t1'om mcmbcrs of councils elected at-larLle. It is clcar lront a few ex-

amples that when state legislators clected from single-melnler districts

133

ack
[me
:ain

;le-
:ild
ier-
ors
in

llv
-e-
nd
;)o
io
.!r-

en
A

JC

le-

3t
':s

lo
rf
.e
re

1S

f-

:S

r€

i-
)r

h
's

3

K



134 Representation and Redistricting lssues

represent parts of cities that use at-large elections, tl're state legislators are
more likely to be asked to help constituents with purely local problems-

In addition to dealing with individual problenrs, the single-member
district legislator is likely to become much more involved in very local prob-
lems: helpirrg citizens get a stopligh[ installed, intervening in zoning ques-
tions, getting state funds to repair a local road or improve a local park. In
other rvords, the state legislator from a single district is more parochial in
his ccncerns. Moreover, the legislator who is conscientious and effective in
handling these local prciblems, and who handles individual problems well,
can build a favorable image that will be valuable in winning reelection.

This brief sutnmary of communications and service patterns is based on
intervielvs with legislators in a number of states with different, sometimes
changing, districting systems. If we are to understand more fully the conse-
quences of districting, we need surveys of constituenrs to find out their
awareness and perceptions of legislators; we also need more detailed ex-
aminations of the service role performed by state legislators-how many
constituenrs are served in rvhat ways and with what results. There are ob-
viously some normative questions involved as well. Should state legislators
be devoting much of their time (and that of their staff) to individual rc-
quests, and should they be focusing attention on parochial projects and
problems? Would the needs of metropolitan communities be better served if
state legislators shared a broader viovpoint? Does at-large repres'entation
actually mean that broader interesw are served or only that conservative
business interests get rnore attention and more direcr representation?
However important these nonnative questions, they cannot be posed ac-
curately-or answered intelligently-un(il we have better empirical answers
to these qucstions about the conscquencss of alternative districting systenr.s"

Consequences for Policy

Probably the most difficult question to ar)swer about districting is the con-
sequences for policy. We may assume that if single-rnember districts pro-
vide more representation for either racial or partisan minorities, there will
be some effects on policy, but measuring these effects is more difficult. We
need to examine how effective such minority representatives are in the
legislature. If representatives front siugle-member districts are less depen-
dent on either party organizations or interest groups for campaign support,
we need to study whether they act independently.

There is some evidence, from roll call Cata, that voting cohesion in stat€
legislative delegations has usually declined when there has been as sHft
from multi- to single-member districts (Jervell, 1969; Harnm, Harmel, and
Thompson, 1979). This is especially likeiy to occur when the change in



Consequences of Districting 135

districting has affected either the llartisan or thc racial balance of the

clisf ricts. Wh.n I intervierved North Carolina legislators' chosen at-large in

metropolitan counties, they assened that this districting system assured

more cohesion among the county delegations on bills and projects affecting

the county. In both Texas and Tennessee, legislators agreed that districting

had reduced the arnount of cohesion on local bills and projects, but there

was considerably more disunity in some metropolitrn delegations than in

others. Divisions within metropolitan delegations resulted from partisan,

racial, and ideological differences, and also from personality conflicts and

political riyalries among the leltislators. Tiiere was also disa-ereement about

whether the decline in legislative dclegation tlarrnony that follorvs single-

member districting is a serious problem or not. Those legislators who shared

the conservative, business-oriented philosophy that used to dominate the

delegations in states like'fexas and Tennessee tended to be critical' Those

whJ represented more liberal viervpoints and racial minorities were not

unhappy about disupity, L'ecattse t|cy considered it inevitable if tl:e view-

points and interests they represented were to be heard in the legislature'

i.low fhat reapportionment has forced state legislatures to pay attention to

metropolitan problems, it seems inevitable that if all the major interests in

the metropolis are goirtg to be represented in the legislature, through

districting, it is going to be more difficult to reach agreenlent on solving

those problems.
The effects of districting on policy making are complex, and they are

not easily measured through aggregate analysis of legislative or budgetary

outputs. What are needed are more detailed studies of the decision-making

pro..r, in legislatures. We need to know how a variety of organized groups

and less organized citizens make their views known to legislators who repre-

sent diffcrint kinds of districts, and how legislators resolve the conflicting

pressures they are subjected to. Is the at'large legislator more vulnerable to

organized group pressures, while the single-member district legislator is

more visible to and thus more sensitive to pressures from individuals and

small groups? We need to knorv more about how recruitment patterns differ

in various districting situati6ns and what effects, if any, this has on policy'

Although it may be difficult to measure poliqy consequences, it seems clear

that districting changes the political environment in which a legislator

operates, the character of policy dcrnancls that are made on him, and the

roles that he must play in responding to those demands'