390-04 Analysis of Ethnic Voting 1997
THE JUNE 1997 VOTE ON WATERFRONT HEIGHT
LIMITATIONS ON MIAMI BEACH:
A BIVARIATE ECOLOGICAL REGRESSION ANALYSIS
OF ETHNIC VOTING
by
Abraham D.Lavender, Ph.D.
and
Chris Girard, Ph.D.
Department of Sociology and Anthropology
Florida International University
Miami, Florida 33199
305-348-2247
September 1997
THE JUNE 1997 VOTE ON WATERFRONT HEIGHT LIMITATIONS ON MIAMI
BEACH: A BIVARIATE ECOLOGICAL REGRESSION ANALYSIS OF ETHNIC
VOTING
by
Abraham D.Lavender, Ph.D.
and
Chris Girard, Ph.D.
Department of Sociology and Anthropology
Florida International University
Miami, Florida 33199
305-348-2247
September 1997
BACKGROUND
Construction of high-rise buildings on the water-front areas
in Miami Beach had become a very controversial issue by 1997 as
increasing numbers of residents objected to these buildings.
Closing of water-front areas to the public, blockage of sun and
breeze to surrounding areas, blockage of water views to the public,
and increasing congestion in driving and parking were among the
objections to the large water-front, buildings. An activist group,
"Save Miami Beach," was formed with David Dermer, a local attorney
and son of a former mayor, as chairman. Sufficient signatures were
collected by "Save Miami Beach" to have a special referendum on
whether or not limitations should be put on large water-front
buildings. On June 3, 1997, a special election was held.
Although there are indications that "Anglo" (non-Hispanic) and
Hispanic voters on Miami Beach are beginning to decrease the
importance of ethnicity in influencing voting behavior, distinctive
differences in voting patterns remain between the two groups. In
November 1995, for example, the last local election held before the
"Save Miami Beach" referendum indicated that in a race for mayor
between two Anglo candidates, the Anglo vote was divided
overwhelmingly with 83.6% for the incumbent candidate (Seymour
Gelber) and 16.4% for the other candidate (Andrew Delaplaine),
whereas the Hispanic vote was almost evenly divided with 54.7% for
Gelber and 45.3% for Delaplaine (Lavender and Girard, January 1996,
p. 9). There also were two races for City Commissioner in November
1995, with each race having one Anglo candidate and one Hispanic
candidate. In one race, the Anglo candidate (Susan Gottlieb)
received 90.8% of the Anglo vote and only 2.62% of the Hispa~ic
vote, while the Hispanic candidate (Ada Llerandi) received 97.4% of
1
the Hispanic vote and only 9.2% of the Anglo vote (p. 12). In the
other race, the Anglo candidate (Marty Shapiro) received 74.2% of
the Anglo vote and 22.8% of the Hispanic vote, while the Hispanic
candidate (Matti Bower) received 77.2% of the Hispanic vote and
25.8% of the Anglo vote (p. 15).
Differences between Anglo voters and Hispanic voters are found
on issues as well as on candidates in the Dade County area. For
example, in Dade County elections in October 1994, heavily African-
American precincts voted 74.3% in favor of a bond issue to provide
scholarships for Florida International University, while the
support for the bond issue was 70.0% in heavily Jewish precincts,
44.9% in heavily Hispanic precincts, and 24.9% in precincts that
were mostly non-Hispanic, non-Black, and non-Jewish. An issue
concerning fire prevention received 74.5% from African-American
precincts, 72.5% from Jewish precincts, 34.8% from Hispanic
precincts, and 49.5% from the other precincts. An issue on a fire
rescue service district received 69.4% from the African-American
precincts, 79.3% from the Jewish precincts, 30.8% from the Hispanic
precincts, and 35.7% from the other precincts. An issue on the
justice system received 58.2% from the African-American precincts,
71.7% from the Jewish precincts, 22.5% from the Hispanic precincts,
and 37.3% from the other precincts (Lavender, November 1994, p. 4).
The ethnic differences noted above in Dade County relate
largely to ideologies on taxing and economic policies, although
modified somewhat by ethnic issues. For example, the Hispanic
support for Florida International University was higher than the
usual Hispanic support for taxing and economic issues, and probably
is related to the perception of Florida International University as
a heavily Hispanic university. On the other hand, the "non-
Hispanic, non-black, non-Jewish" precincts gave less support to the
FlU bond issue than to the other issues, probably because of the
same reasons that influenced Hispanic voters to be relatively more
supportive. Ethnic differences also include issues on which the
average voter might not expect a difference. In the November 1995
election in Miami Beach, for example, there also was an issue as to
whether or not a park in the South Beach area should be renamed for
Mrs. Marjory Stoneman Douglas. Anglo voters voted 62.5% to name the
park after Mrs. Douglas, and 37.5% against the name change. While
Anglo voters went two-to-one for the name change, Hispanic voters
had the opposite outcome. Hispanic voters voted only 30.7% for the
name change, and 69.3% against the name change.
In the week before the "Save Miami Beach" vote on June 3,
1997, informal and impressionistic observations suggested that this
election also might have an ethnic factor. The impression was that
the Anglo community would support the building limitations, but
that the support was very low among Hispanic voters. Those who
opposed the building limitations had a very large monetary account,
and had advertised heavily in the Hispanic community as well as in
the non-Hispanic community. Hence, in the last few days of the
campaign special attention was given to the Hispanic community by
those who supported the building limitations. In addition to the
usual issues mentioned above, it was noted that the major developer
2
who was most strongly associated with building highrises on the
waterfront (Thomas Kramer) had also "suggested Miami would lose its
slums when Cuban exiles moved back to Cuba. And he spoke about New
York, where 'the quality of the immigrants was reduced from smart,
wealthy people to Puerto Ricans'" (Balmaseda, May 25, 1994). Heavy
attention was given to the possible relation between highrise
waterfront development and its effect on displacement of the
Hispanic community. This article analyzes the ethnic factor in the
"Save Miami Beach" vote.
VOTER REGISTRATION AND TURNOUT
At the time of the election, Miami Beach had 39,330 registered
voters, and 10,931 ballots were cast on June 3, 1997 for a turnout
of 27.8%. Of the 10,931 ballots cast on June 3, 1997, 10,707 were
counted as valid. The limitation on building waterfront highrises
passed with 6,145 votes (59.4%) in favor, and with 4,562 votes
(42.6%) opposed to limitations.
In the November 1995 elections, when there were elections for a
mayor, two members of the City Commission, and the park name change
issue, only 30.7% of the registered voters cast ballots.
Considering that only one issue was on the ballot on June 3, 1997,
the turnout of 27.8% suggests a relatively high level on this
issue. As noted below, when using the registration and voting
figures after the registration roll was purged within the few weeks
after the election, the turnout rate increases to 28.1%.
In February 1991, 72.1% of the registered voters in Miami
Beach were non-Hispanic, and 27.9% were Hispanic. In January 1993,
the non-Hispanic voters had decreased to 69.0%, and the Hispanic
voters had increased to 31.0%. In November 1995, the non-Hispanic
voters had decreased further to 62.8%, and the Hispanic voters had
increased further to 37.2% (Lavender and Girard, January 1996, p.
3). Hispanic turnout generally is higher than the non-Hispanic
turnout, and it was suggested that 42.6% of the ballots cast had
been cast by Hispanic voters in November 1995 (Lavender and Girard,
p. 9). The changing percentages of Hispanic and non-Hispanic
voters, the higher turnout by Hispanic voters, and the fact that
there also was an ethnic component to the June 3, 1997, election
because of Kramer's comments all combined to emphasize the
importance of analyzing the election by ethnic groups.
The Dade County Board of Elections does not "freeze" its data
on the day of an election. Names are being added and removed on a
daily basis. In the few weeks after an election that it takes to
get the results of which individuals did and did not vote, a very
small percentage of changes have been made. Hence, the numbers will
not match exactly. In addition, shortly after the election on June
3, 1997, and before data could be obtained from the Board of
Elections, the voting rolls were purged and updated. As a result of
these additions and deletions, analysis of the June 3 vote is based
on 38,438 registered voters rather than on 39,330 registered
3
voters. Even with the purge, the decrease in number of voters was
only 2.3%. Voters who are purged are almost always purged because
they have not voted in a long time. There possibly were individuals
who voted on June 3, and either died or moved away immediately
after the June 3, but very few of these would be among those
removed from the rolls in the short time between the June 3 vote
and the purging in the following few weeks. A small number of
voters probably did still live in Miami Beach, but were purged
because they had not voted in a number of years. The vast majority
of those purged, however, probably are deceased or have moved out
of Miami Beach. Hence, the list of 38,438 individuals on the
registration rolls after the purge probably is more indicative of
those people who really were eligible to vote on June 3. That is,
we probably can assume that almost none of those people who were
purged would have been eligible to vote June 3, either due to death
or moving away before that date. Using the updated list of 38,438
registered voters, analyzes indicate that 40.6% of the registered
voters were Hispanic, but that, because of the higher turnout,
Hispanic voters cast 47.0% of the total ballots.
As expected, the turnout rate varied tremendously for
different ethnic or nationality groups, as shown in Table 1. Cuban-
born voters had by far the highest turnout, 46.1%. In contrast,
Puerto Rican-born voters had only a 22.2% turnout. Overall,
Hispanic voters had a 32.5% turnout rate, much higher than the
25.0% turnout rate for non-Hispanics. It was this higher turnout
rate that led to Hispanics casting, as noted above, 47.0% of the
total votes while comprising 40.4% of the registered voters.
There also was a small difference between Jewish voters and
non-Jewish, non-Hispanic ("Other") voters. The Jewish voters turned
out at 27.1%, while the "Other" voters turned out at 22.3%.
In Table 2, the ethnic groups are further divided by three age
groups (18-40, 41-64, and 65+) . As shown in Table 2, there also was
a strong relation between age and voter turnout rates, with the
young voters having a 15.5% turnout, the middle-aged voters having
a 31.3% turnout, and the elderly voters having a 35.6% turnout.
When age was combined with ethnicity, there were very large
differences in turnout rates. Elderly Cuban-born voters had the
highest turnout, 54.8%. In contrast, for example, young Cuban-born
Hispanics had a 21.8% turnout, and young United States-born
Hispanics had only a 13.6% turnout. For all Hispanic voters, the
turnout rate was 32.5%, over two times the rate (14.0%) for young
Hispanic voters.
The age pattern was found for all ethnic groups, regardless of
the ethnic group's overall turnout rate. Middle-aged Jewish voters,
for example, had a 32.2% turnout rate, but young Jewish voters had
only a 13.9% turnout rate. Elderly Jewish voters had a turnout rate
slightly lower than the turnout rate for middle-aged Jewish voters.
"Other" voters had the same turnout pattern as Jewish voters had,
with the youth turnout being 17.1%, the middle-aged having the
highest percent, 29.6%, and the elderly being in between the youth
and the middle-aged with 21.4%. It is noted that all non-Hispanic
voters, including both Jewish voters and "Other" voters haa a
4
turnout of 24.8%, compared to a Hispanic turnout of 32.5%. However,
within the non-Hispanic voters, the Jewish voters had a higher
turnout (27.1%) than the "Others" (22.3%). The low turnout of all
"Other" voters was largely because of the general tendency for
young voters to have low turnouts, and the fact that young voters
account for a much higher percentage of the "Other" voters than
they do for either Jewish or Hispanic voters. It is noted that the
"Other" young voters had a higher turnout (17.1%) than either the
Jewish young voters (13.9%) or the Hispanic young voters (14.0%).
5
Table 1. Voter Turnout by Ethnic Subgroups
Ethnic or
Nationality
Group
Number of
Registered
Voters
Number
of
Voters
Percent
Turn-
out
Foreign-born Hispanics 8,919 3,591 40.3
Cuban-born 6,225 2,868 46.1
Non-Cuban 2,694 723 26.8
Puerto Rican 827 184 22.2
Colombian 484 143 29.5
Hispanic Self-Identity* 4,742 1,122 23.7
U.S.-Born Hispanic Sample 1,900 352 18.7
Jewish Sample 12,004 3,252 27.1
"Other" Sample** 11,041 2,467 22.3
Total non-Hispanics 22,877 5,722 25.0
Total Hispanics 15,561 5,065 32.5
Total Voters*** 38,438 10,787 28.1
*Those voters who registered for the first time in January 1995 or
later, and self-identified themselves as Hispanic byethnicity, but
did not have a space in which to list place of birth because the
voter registration form was changed effective January 1995.
**Non-Hispanic and non-Jewish.
***There area about 200 voters who are both Jewish and Hispanic.
6
Table 2. Voter Turnout by Ethnic and Age Subgroups*
Ethnic Subgroups
Foreign-born Hispanics
(N=8,919)
Cuban-born
(N=6,225)
Non-Cuban born
(N=2,694)
Puerto Rican-born
(N=827)
colombian-born
(N=484)
Hispanic Self-Identity*
(N=4,742)
U.S.-Born Hispanic Sample
(Sample=296i est. N=l,927)
Jewish Sample
(Sample=886i est. N=12,004)
"Other" Sample**
(Estimated N=11,041)
Total Hispanics
(Estimated N=15,588)
Total non-Hispanics
(Estimated N=23,045)
Age Groupings of Registered Voters
18-40 41-64 65+ Total
21.8
(1317)
26.9
(662)
16.6
(655)
12.7
(259)
22.2
(99)
9.1
(1968)
13.6
(1471)
13.9
(1856)
17.1
(5314)
14.0
(4756)
16.6
(7170)
34.2
(3544)
38.7
(2230)
26.6
(1314)
23.4
(380)
28.4
(257)
26.3
(1549)
29.8
(312)
32.2
(2696)
29.6
(4018)
31.6
(5405)
30.7
(6714)
51.6
(4058)
54.8
(3333)
36.6
(725)
33.0
(188)
37.5
(128)
43.8
(1225)
45.5
(144)
28.5
(7452)
21.4
(1709)
49.6
(5427)
27.3
(9161)
40.3
46.1
26.8
22.2
29.5
23.7
18.7
27.1
22.3
32.5
24.8
-----------------------------~------------------------------------
Total Voters
(N=38,434)*
15.5
(11830)
31.3
(12047)
35.6
(14557)
*A few voters are loss due to lack of information.
7
28.1
BIVARIATE ECOLOGICAL REGRESSION ANALYSIS
Having shown that there were different rates of turnout
between non-Hispanic and Hispanic voters, it is important to
determine if there was a difference in support for the "Save Miami
Beach" vote. As shown in Figure 1, there was a significant
difference in voting for or against the building limitations. As
expected, the non-Hispanic voters supported the limitations much
more than did the Hispanic voters. Non-Hispanic voters voted for
the building limitation by 82.7% to 17.3%. Although it had been
assumed that the building limitations would receive very little
support from the Hispanic community because of the major media
blitz opposing the building limitations, Hispanic voters cast 31.7%
of their votes for the building limitations and 68.3% against the
building limitations. These numbers include absentee voters.
8
Figure 1. Bivariate Ecological Regression Analysis
75
100
50
25
Y
V
0
T
0
0 25 50 75 100
HISPVOTE
Plot of YVOTE with NONHVOTE
100
75
50
y 25
V
0
T
E 0
0 25 50 75 100
NONHVOTE
REFERENCES
Balmaseda, Liz. "Beach Paint Job Can' t Camouflage Developer's
Colors." The Miami Herald, May 25, 1994.
Lavender, Abraham D., and Chris Girard. "The Miami Beach Mayor and
Commissioners Election of November 1995: An Ethnic Analysis."
Florida International University, January 1996.
Lavender, Abraham D. "The FlU Scholarship Vote of October 1994: An
Analysis of Race, Ethnicity, Social Characteristics, and Region of
Residence in Dade County." Florida International University,
November 1994.
This analysis was conducted with the cooperation of the "Save Miami
Beach" Political Action Committee. Appreciation is expressed to
Jose prendes of the Computer Science Center at Florida
International University, North Campus.
10