Racially Polarized Voting in Jefferson Parish, Louisiana by Richard Engstrom
Working File
January 1, 1988
Cite this item
-
Case Files, Chisom Hardbacks. Racially Polarized Voting in Jefferson Parish, Louisiana by Richard Engstrom, 1988. 2d7bb225-f211-ef11-9f8a-6045bddc4804. LDF Archives, Thurgood Marshall Institute. https://ldfrecollection.org/archives/archives-search/archives-item/5a07ecc3-d7ef-45b4-91db-37aa3827e50e/racially-polarized-voting-in-jefferson-parish-louisiana-by-richard-engstrom. Accessed November 28, 2025.
Copied!
RACIALLY POLARIZED VOTING
in JEFFERSON PARISH, LOUISIANA
By
RICHARD L. ENGSTROM, PH.D.
RESEARCH PROFESSOR of POLITICAL SCIENCE
UNIVERSITY OF NEW ORLEANS
On 20 occasions since 1980, voters within Jefferson
Parish, Louisiana, have been presented with a choice between or
among black and white candidates
the Parish. Analyses of these
below, reveal that if only the
these elections had been counted,
been elected in 14 or 15 of them.
the votes cast by whites had been
candidate would have been elected.
seeking public offices within
elections, reported in detail
votes cast by black people in
black candidates would have
If, on the other hand, only
counted, not a single black
When black candidates have
competed with white candidates in Jefferson Parish in the 1980's,
there has invariably been a consistent relationship between the
racial composition of the precinct electorates and the vote cast
within those precincts, and the magnitude of the estimated
differences in the candidate preferences of white and black
voters has usually been .large. Voting in the vast majority of
these elections, therefore, has been "racially polarized*, as
defined by the United States Supreme Court in Thornburg V.
ingles,, 106 S.Ct. - 2752, at 2768 n.21 (1986). Indeed, in 10 of
these
black
elections, the estimated racial division in the vote for
candidates exceeds the largest division in the vote for any
election within the invalidated multi-member legislative
districts reported in Thornburg (at 2782 - 2783).
The 20 elections have been analyzed by employing both
bivarate ecological regression analysis and the complementary
technique of extreme case (on homogeneous precinct) analysis.
2
These analytic procedures have been applied, as in Thornburg, in
order to derive 'estimates of the percentage of members of each
race who voted for black candidates' (at 2768). More
specifically, both the regression analysis and the extreme case
analysis have been performed in a manner that produces estimates
of the 'percentage of candidates"(Thornburg, at 2782 - 2783).
The following reports the estimated racial divisions in
the vote for the black candidates in each of these elections. In
order to simplify the presentation, estimates derived through
regression will be presented first, followed parenthetically by
those based on voting within homogeneous precincts. Estimates
from regression are based on comparisons between the race of the
voters and the support for the black candidates in every precinct
in which votes have been cast, while those from homogeneous
precincts are based only on precincts that are predominantly
black or predominantly white (i.e., precincts in which at least
90% of the electorate is of one race)
In addition to the estimated divisions in the vote, the
actual 'correlation between the racial composition of precincts
and support for the black candidates in each precinct, as
measured by the 'correlation coefficient' or Pierson's r, is also
reported for each election. This coefficient measures how
consistently, over all of the precincts, the vote for the black
candidate varies with the racial composition of the precincts.
3
The maximum value that this coefficient can reach is
.1.0. The lowest value for r obtained for-any of these elections
was .759. In the vast majority of elections, the vote for the
black candidate correlated with the racial composition of the
precincts at a level greater than 9.0.
Data reflecting the race of the people who signed in
the vote in the various precincts on the day of the election have
been provided by the Register of Voters in Jefferson Parish for
the 15 elections held prior to August 30, 1986, and these data
serve as the measure of the racial composition of the precincts
for those elections. Sign-in data are not available, however,
eor elections held on that date or later, and therefore, 'the
racial composition of the precincts in these elections is based
on the race of the people registered to vote in the election
within the precinct. The vote within each precinct is taken from
the official election returns.
• I.
PARISHWIDE
Only one of the 20 elections involves a black candidate
seeking a parishwide office in Jefferson. In August of 1987, Mr.
Anderson Council, a black, was a candidate for Judge of Juvenile
Court, Section A. The correlation between the race of the
registered voters in this election and, the vote for Mr. Council
4
across the 255 precincts in Jefferson was very high, .914. Mr.
Council, who competed with. two white candidates for this
position, is estimated to have received 80.8% (74.9%) of the
votes cast by blacks, but only 3.1% (4.5%) of those cast by
whites, and as a consequence finished last among the three
candidates.
KENNER
The City of Kenner, located in the western portion of
the East Bank of Jefferson Parish (see Map A), is the largest
municipality in Jefferson. Black candidates have competed with
white candidates for municipal office in Kenner on five occasions
since 1980, three of which involved elections to citywide offices
in this majority white city. The latest citywide candidacy was
that of Emmitt N. Richardson, who sought an at-large seat to the
Kenner City Council in a special election in May of this year.
The correlation coefficient between the race of the registered .
voters and the vote for Mr. Richardson across the 31 precincts in
Kenner was .988. Despite competing with five white candidates,
Mr. Richarson is estimated to have received 80.1% (80.9%) of the
votes ca.st by blacks, but virtually none (4.0% in the
homogeneous precincts) of the votes cast by whites, and he
therefore finished fourth among the six candidates.
Two black candidates competed with white candidates for
citywide posts in April of 1986. Michael Austin, a black,
opposed Salvador J. Lentini for the position of Chief of Police.
5
The correlation coefficient between the race of the people who
signed-in to vote that day and the vote for Mr. Austin across the
31 precincts was .903. Support among black voters for these two
candidates was virtually even. Mr. Austin is estimated through
regression to have received 49.1% of the votes cast by blacks,
and 50.1% through on examination of the votes cast in blAck
homogeneous precincts. White voters strongly supported the white
candidate, however, casting only 14.7% (15.3%) of their votes for
Mr. Austin, and Austin was therefore soundly defeated.
The other contest in April, 1986, was for the election
at-large of two members of the city council. Oscar Code, a
black, sought an at-large seat along with two whites, Nicholas J.
Baroni and Edmond J. Muniz. In this election, every voter was
permitted , but not required to cast votes for two candidates.
When the vote for the different candidates is regressed onto the
percentage of voters who signed-in that day, Mr. Cade is revealed
to be the second choice among black voters, but the last choice
of the white voters. Mr. Baroni is estimated to have received a
vote from 19.7% (53.0%) of the blacks who signed-in, Mr. Cade,
42.6% (40.0%), and*Mr. Muniz 32.9% (37.0%). Mr Cade, however,
received votes from only 12.1% (13.07%) of the whites who signed
in, and therefore, both of the seats were awarded to the white
candidates.
Another recent election in Kenner involving black and
white candidates was the 1982 election for the District 1 seat on
6
the city council (See Map B). At the time of this election, a
majority (54.8%) of the registered voters in District 1, which
covers 7 precincts in Kenner, was black. Three black candidates,
Wilma K. Irvin, Shirley G. Burton, and Anderson Council sought
this seat along with two white candidates, John M. Stoulig and
Harry Polito. The votes for each of the black candidates
strongly correlated with the racial composition of the precincts
(Ms. Irvin, r = .963; Ms. Burton, r = .921, and Mr. Council, r =
879). The black candidates collectively received 95.8% (97.3%)
of the votes cast by blacks, distributed as follows; Ms. Irvin,
39.5% (36.4%), Ms. Burton, 32.4% (28.6%), and Mr. Council, 23.9%
(32.3%). Ms. Irvin, in turn, is estimated to have received only
8.02% (7.9%) of the votes cast by whites, Ms. Burton, 8.8%
(8.6%), and Mr. Council 0.0% (2.4%). Blacks constituted 56.3% of
the voters participating in this election, and their most
preferred candidate, Ms. Irving, was placed in a runoff with Mr.
Stoulig. In the runoff, 56.4% of the people voting were black.
Ms. Irvin's vote .correlated at .984 with the racial composition
of the precincts. She is estimated to have received 82.2%
(88.0%) of the votes cast by blacks along with 11.9% (15.9%) of
those cast by whites, which was sufficient to elect Ms. Irvin.
School Board, Districts 3 and 7
District 7 of the Jefferson Parish School Board
encompasses 26 precincts on both the East and West Banks of
Jefferson Parish (See Map C). In the September, 1984 election,
Herbert Wallace, Jr., a black, was a candidate for this seat on
7
the board along with two white candidates. Mr. Wallace's suppo'rt
across the 26 precincts was correlated with the racial
composition of the precinct electorates at .759. Blacks voting
in this election are estimated to have cast a majority of their
votes, 57.5% (61.9%) for Mr. Wallace. Whites, however, gave Mr.
Wallace only 8.7% (10.8%) of their votes, and he therefore was
defeated.
A second school board seat, that for District 3, had
been contested the previous year by Ellis Wilson, a black.
District 3 encompasses 21 precincts on the West Bank east of
District 7 (Map c). Six white candidates also sought the
district 3 seat in the October, 1983 election. Mr. Wilson's vote
correlated with the racial composition of the precinct at .995.
He is estimated to have received only 3.2% (5.2%) of the ,votes
cast by whites, but 84.2% (8.4.4%) of those cast by blacks. The
white vote was severely divided across the six white candidates,
resulting in Mr. w4lson placing first in this election, 95 votes
ahead of the second place candidate, Robert A. .Autin. In the
run-off election in November, Mr. Autin soundly defeated Mr.
Wilson. The vote for Wilson again correlated strongly, .981,
with the racial composition of the precincts. His share of the
black votes increased to 93.5% (93.0%) in this election, while
his share of the white vote is estimated to have been 18.5%
(20.8%)
8
7th Justice Court
Located between school board District 3 and 7 on the
West Bank (and overlapping District 7 slightly ) is the
geographical area for the 7th Justice court (see Map D). This is
a recently created majority black district that encompasses 5
precincts. A special election to choose a Justice of the Peace
and Constable in this district was held on April 4th of this
year. At the time of this special election, 66.1% of the
registered voters in the District was black. Mr. Hebert Wallace,
a black, successfully sought the position of Constable, soundly
defeating a single white opponent, Mr. Bryan Hymel. Mr.
Wallace's vote correlated with the black percentage of
registration in the precincts at .881. The turnout for this
constable election, however, was very low, as only 25.8% of the
registered voters cast a ballot.
Turnout was especially low among whites. Only 11.9%
(13.2%) of the white's registered -voters voted for a constable
candidate, compared to 32.8% (35.4%) of the blacks. Given the
small 'number of precincts and -low white turnout, precise
estimates of the racial division in the vote must be interpreted
with caution. Both regression and homogeneous precinct analy'ses
provide similar estimates of Mr. Wallace's share of the black
vote, 94.5% and 93.6% respectively, and there should be little
doubt that he was the preferred candidate among black voters.
Estimates of this share of the white vote differ substantially,
however, from 0.0% (regression) to 31.8% (in the single white
9
homogeneous precinct). Even accepting the more generous estimate
of 31.8%, Mr. Wallace can hardly be considered the choice of the
white voters, especially in light of the fact that Mr. Hymel won
substantial majorities (68.2% and 64.9%) in each of the majority
white precincts.
State House District 87
Thirteen precincts in the Marrero and Harvey area of
the West Bank
Representatives
candidate for
comprise District 87 of the Louisiana
(see Map E). A black, Lloyd Lewis,
this legislature seat along with five
House of
Sr. was a
whites in
October 22, 1983 election. The correlation between the racial
composition of the election in these 13 precincts and the vote
for Mr. Lewis was .998. He was the choice of over a majority of
the blacks participating in this election, receiving 60.8%
(60.0%) of their votes. White voters, however, gave him little
support. He is estimated through regression to have received
0,0% of the votes cast by whites, and through homogeneous
precincts, only 2.2%. The white vote was severely dispersed
across the five white candidates however, which allowed Mr. Lewis
to finish second (by 10 votes) and earned a position in a runoff
election with N.J. Damico. In the runoff, the vote for Mr. Lewis
again correlated strongly with the racial composition of the
precincts, .998, and his support among black voters is estimated
to have risen to 94.3% (93.9%). This was not sufficient to elect
Mr. Lewis, however, as he received the votes of only 5.8% (7.7%)
of the whites voting.
10
Gretna
Located in the eastern end of the West Bank of
Jefferson Parish is the City of Gretna (Map F) . Voting in
Gretna was recently found to be racially polarized in Citizens
for a Better Gretna V. City, of Gretna, Civ. No. 84-4901,
(U.S.D.C., E.D. La. May 12, 1986). Since that litigation, blacks
on two occasions have competed with white candidates for
municipal office in Gretna. In the March 1985 election, Roy
Wilson, a black, was a candidate for Chief of Police, along with
two whites, Beauregard H. Miller, Jr. and Joseph F. Cerniglia.
The vote for Mr. Wilson in this election correlated with the
racial composition of the 13 precincts in Gretna at .974. He is
estimated, through regression, to have received 64.7% of the
votes cast by blacks (no black homogeneous precincts exist in
Gretna on which to base a second estimate). His vote among
whites is estimated to have been only 1.3% (3.2%), however, and
he was, therefore, defeated.
In April of this year, Gretna held an election for the
seats on its city council, employing the new electoral
arrangement adopted following the Citizens for a Better Gretna
decision. A majority black district in that new arrangement,
District 11 was contested only by black candidates. A second
district, No. 3, was also contested by a black, Terry Videau,
along with seven white'candidates. District 3 comprises only tow
precincts (map G), neither of which is racially homogeneous, and
therefore, no statistical coefficients will be reported for this
election.
11
It can be noted, however, that as in other elections involving
black 'candidates in Gretna, Mr. Videau's vote was correlated with
the racial composition of the precincts, in that he received
16.7% of the votes in the precinct in which 27.3% of the
registered voters was black, and 21.1% in the precinct in which
blacks comprised 35.6% of the registered voters. Mr. Videau was
able to finish second among this large field of candidates, and
therefore competed in a runoff election in may. In the runoff,
Mr. videau's vote was even more closely linked to the racial
composAion of the precincts; he received 26.3% of the votes in
the precinct in which voter registration was 28.0% black, and
31.6% of the votes in the precinct that was 35.9% black in
registration. This close association between the vote for Videau
and the presence of black registered voters, in light of the
usual racial divisions in the support for black candidates in
Gretna, suggests that Mr. Videau was the preferred candidate of
the black voters in District 3.
Other Black Candidates
Black Voters in Jefferson Parish have not given their
support to every black candidate in the Parish. On five
occasions .since 1980,, the estimated divisions in the vote allow
one to infer that the black candidate would have lost the
election even if only the votes cast by blacks had been counted.
Shirley Burton, for example, has twice been a candidate for the
Louisiana House of Representatives from District 92 with the
Parish, in 1983 and 1986. In the first attempt, she is estimated
12
•
to have received 45.7% (41.3%) of the black vote and 15.9%
(16.56) of the white. In the second, she received 46.1% (37.9%)
and 3.8% (5.4%), respectively. In 1983, two blacks were
candidates for the parish council. Mr. Leon T. Williams sought
the councilman-at-large position for Districts 3 and 4, receiving
28.5% (31.4%) of the votes cost by blacks and 4.5% (4.7%) of
those cost by whites, Mr. Robert Hamilton sought the District 2
seat, receiving an estimated 27.8% (27.1%) of the votes of black
and 3.9% (4.6%) of those of whites. Finally, Sylvester Brown, a
candidate for the District 1 seat on the parish school board in
1984, received an estimated 29.1% of the votes cast by blacks
(there were no black homogeneous precincts in the district), and
0.5% .(0.9%) of those cast by whites.
Black voters in Jefferson Parish have demonstrated, on
numerous occasions since 1980, a preference for black candidates
over white candidates. When this has happened, the preference
for the black candidate has never been shared by the white
voters. Only two of these black candidates (Mr. Irvin and Mr.
Wallace) have been elected, each in a majority black electoral
unit (District 1 in Kenner and the 7th Justice District,
respectively). Among tbe others, three have progressed as far as
a runoff election, only to then lose to a white candidate.
Voting in Jefferson Parish is clearly racially polarized, to a
degree that has usually precluded black voters from having an
equal opportunity to elect the candidates of their choice.
13
1. See, Grofman, Migalski, and Nouiello, the
°Totality of Ciraumstances Test° in Section 2 of the 1982
Extension of the Voting Rights Act: A Social Science Perspective,
7 LAW AND POLICY 119, 202 - 205 (cited in Thornburg V. Gingles,,
106 S.Ct. 2752, at 2768 (1986). This methodology is also
explained in Grofman, An Outline for Racial Bloc Voting Analysis,
Plaintiffs' Exhibit 12, at 1-5 1 Gingles V. Edmisten, 590 F.Supp.
345 (1984).
2. These estimates are also reported in tabular form
in Appendices A and B,
estimates reported in the
are derived from weighted
respectively. The regression-based
text of this report and in Appendix A
regressions that take into account the
vast differences in.the relative size of precinct electorates.
Estimates based on unweighted regressions, in which every
precinct is treated as an equal unit regardless of the size of
the vote within it, are reported in Appendix C. The conclusions
expressed in this report are supported by the results of both the
weighted and the unweighted analysis.
3. Th9 correlation coefficient can be interpreted as
a measure of how well the data fit the
which ecological regression is based.
if r > .50, then the 'fit' should be
assumption of linearity on
Grofman has suggested that
considered adequate. See
14
•
Grofman, Migalski, and Noviello, supra. note f, at 206. John K.
Wildgen has further noted that a correlation coefficient of .867,
in the context of a voting analysis such as this, is 'remarkably
good'. See Wildgen, A Political Problem of Public Education' in
New Orleans, 8 LA. BUS. SURVEY, 5, 6n.3 (July, 1977).
4. The race of the people signing-in to vote has been
recorded by the Registrar of Voters as either black or white. A
small percentage of the registered voters however, (currently
0.5%) has been classified as "other". When relying on
registration data to measure the racial composition of precincts,
the percentage of registered voters who are black serves as the
measure. This measure therefore, results, technically, in
estimates of the voting behavior of blacks and non-blacks. Give
that virtually all of the non-blacks are white (currently,
99.5%), the estimates for non-blacks will be reported in the text
as estimates for whites.
15
•
APPENDIX A
Results of Weighted Regression Analyses
Z of white Zof black .
Black voters for black voters for black Correlation
Election Year Candidate Candidate Candidate Coefficient
Juvenile Court,
Section A 1987 Council 3.1 80.8 .914
Kenner Council,
At-Large 1987 Richardson 0.0 80.1 .988
Kenner Chief
of Police 1986 Austin 14.7 49.1 .903
Kenner Coucil
At-Large 1986 Cade 12.1* 42.6* .940
Kenner Council
District 1 1982 Irvin 8.0 39.5 .963
Burton 8.8 32.4 .921
Council 0.0 23.9 .879
Kenner Council
Dist. 1 Runoff 1982 Irvin 11.9 82.2 .984
School Board,
District 7 1984 Wallace 8.7 57.5 .757
School Board,
District 3 1983 _Wilson 3.2 84.2 .995
School Board, •
Dist.3 Runoff 1983 Wilson 18.5 93.5 .981
Constable,
7th Jus.Dist., 1987 Wallace 0.0 94.5 .881
House Dist.87 1983 - Lewis 5.8 94.3 .998
House Dist.87
Runoff 1983 Lewis 5.8 94.3 .998
Gretna Chief
of Police 1985 Wilson 1.3 64.7 .974
Figure R for the percentage of whites and blacks
signing-in or election day, not the percentage
voting in the election for the at-large seats.
16 •
APPENDIX B
Results of Homogeneous Precinct
2 of white 2 of black
Black voters for black voters for black
Election Year Candidate Candidate Candidate,
Juvenile Court,
Section A 1987 Council 4.5 74.9
Kenner Council,
At-Large 1987 Richardson 4.0 80.9
Kenner Chief
of Police 1986 Austin 15.3 50.1
Kenner Coucil
At-Large 1986 Cade 13.0* 40.0*
Kenner Council
District 1 1982 Irvin 7.9 36.4
Burton 8.6 28.6
Council 2.4 32.3
Kenner Council
Dist. 1 Runoff 1982 Irvin 15.9 88.0
School Board,
District 7 1984 Wallace 10.8 61.9
School Board,
District 3 1983 Wilson 5.2 84.4
School Board, •
Dist.3 Runoff 1983 Wilson 20.8 93.0
Constable,
7th Jus.Dist. , 1987 Wallace 31.8 93.0
House Dist.87 1983 - Lewis 2.2 60.2
House Dist.87
Runoff 1983 Lewis 7.7 93.9
Gretna Chief
of Police 1985 Wilson 3.2 NA
Figure B for the percentage of whites and blacks
signing-in or election day, not the percentage
voting in the election for the at-large seats.
17
•
APPENDIX C
Election.
Results of unweighted Regression Analyses
.2 of white 2 of black
Black voters for black voters for black Correlation
Year Candidate Candidate Candidate Coefficient
Juvenile Court,
Section A 1987 Council
Kenner Council,
At-Large 1987 Richardson
Kenner Chief
of Police 1986 Austin
Kenner Coucil
At-Large 1986 Cade
Kenner Council
District 1 1982 Irvin
Burton
Council
Kenner Council
Dist. 1 Runoff 1982 Irvin
School Board,
District 7 1984 Wallace
School Board,
District 3 • 1983 Wilson
School Board,
Dist.3 Runoff 1983 Wilson
Constable,
7th Jus.Dist. , 1987 Wallace
House Dist.87 1983 Lewis
House Dist.87
Runoff 1983 Lewis
Gretna Chief
of Police 1985 Wilson
3.1
0.0
14. 7
11.9*
10.0
9.1
77.2 .909
78,3 .988
48.4 .871
41.5* .917
38.5 .949
31.7 .937
0.0 24.5 .905
15.1 81.9 .979
8.4 55.3 .753
2.7 84.2 .994
18.4 .93.8 .975
0.8 92.7 .894
0.0 60.5 .998
5.9 94.5 .998
3.2 NA .981
Figure B for the percentage of whites and blacks
signing-in or election day, not the percentage
voting in the election for the at-large seats.
18
I.
•
MAP A
CITY OF KENNER
z
•
* • • JEFFERSON PARISH. LA. • • *
WARD & PRECINCT MAP
*Aft* ockammings
Pea CasC 1 1110.0.0.6All •
1111,•••• 0/0.1.118.1 IMAMS
•;,•.
-
•
4ttrf4r.
• •
vvvvvvv •••,.•
1.1•••4•• ..•1
emboaft. to .11.11*
SIM ONO ••••Mlo
111.MM•
•
MAP B
DISTRICT 1
CITY OF KENNER
tilts:01
.4 %1
Pi rilikitiV7:°1/4)6-.1(111111111111111s1
Iril 11 ' lill*Ifiliiii 111121111111111111111k ;la
11
it" CI i ? Illaimmit\IRPA
It 111111111111=rmila 1M
- Ill II rt11111111121-A m 1111111111111 N
uniniattga4mt
i iirit4' %., Um* mitiumwtutif
..41114=1 % lizokre Rita; i , ittremetura ii „Ty II tt r.r.orr.-4103 110-.0-iiiiimilyii litioniinEumt
oic ogotiffitimill
Intrittlinnaitif ° Frrilgra
1 iiiii I li 711; 111:11.41:111:11:111 ritilit:1111111 1
...a,....0.0
mmeroshild
ill
1-41 111111
O
si ......fibt
:amp
trant110111:17111411171
mini- --4 ----imPimedatv
,
.z.•
MAP 0
.7thJUSTICE DISTRICT
• •
MAP E
DISTRICT 87
•LOUISIANA HOUSE OF REPRESENTATIVES
;•/.q
\,\ I.
* * * JEFFERSON PARISH . LA.* * *
WARD IL PRECINCT MAP
•Aso
• Olive OCVMDAPIIII •
0
ONO • • ••••••••
•
• 'no.
41•1•11.1%. •■•••••••• MAIM.
ol • • • ••••
••• • • • • 11 •
01.1.01.11. O..
Owe NM OF .11.1/10
MAP F
CITY OF GRETNA
S
MAP G
CITY OF GRETNA
•
1:‘