A Natural Experiment of Race-Based and Issue Voting: The 2001 City of Los Angeles Elections

Published date01 June 2005
Date01 June 2005
Subject MatterArticles
A Natural Experiment of Race-Based and Issue Voting:
The 2001 City of Los Angeles Elections

The theory of racially polarized voting suggests that race is a primary determinant of vote choice in elections
where a minority candidate is pitted against a white candidate. The spatial model of voting suggests that voters
consider the issue positions of candidates and choose the candidate closest to their own positions. The unique
context of the 2001 Los Angeles city election allows us to test these two theories. In each of two races in this
election, a Latino candidate competed against a white candidate. In one race the white candidate was consid-
ered more liberal, while in the other race the Latino candidate was seen as more liberal. This particular ethnic
and ideological composition provides us with a natural experimenting which to test the two competing theo-
ries. While voter ethnicity mattered, we show that consistent with the spatial model, voters also relied on issues
and ideology as factors in their voting choices. By considering the choices voters are making in two different
elections, we argue that estimates of the extent of racial voting in previous research may be overstated.
Most research on political behavior has focused on equal in these two races, yet the amount spent in the city
four major theories of voter decisionmaking:
attorney election is considerable, and these spending figures
sociodemographic characteristics of voters (e.g.,
indicate that both elections were competitive.2 Second, not
Berelson and Lazarsfeld 1944), partisanship (e.g., Campbell
only were both of these positions open-seat races, but Los
et al. 1964), issues and ideology (e.g., Key 1966), and eco-
Angeles city elections are nonpartisan, and the candidates
nomic conditions (e.g., Key 1966). Much of this literature
competing in the general elections were Democrats. Thus,
debates the relative importance of these characteristics, with
the cue of partisanship was absent from these elections
recent research focusing on issues and economic factors
(Schaner and Streb 2002; Squire 1988), as well as its next
(e.g., Alvarez and Nagler 1995).
best substitute, incumbency (Schaffner, Streb, and Wright
In this article we focus on the 2001 Los Angeles citywide
2001). This setting gives us the ability to understand how
general elections. Both the 2001 Los Angeles mayoral and
voters decide in this setting.
city attorney races presented a unique opportunity for
Another important reason to study the 2001 Los Angeles
researchers of voting behavior. First, these two simultaneous
city elections is that the races featured candidates of differ-
elections, both for executive-style city government posi-
ent ethnic identities, as a Latino candidate competed against
tions, provide more information for studying voter deci-
a white for each of the two highest elected offices in city
sionmaking strategies than does a single election; also, by
government. The ethnic context of this election produces a
looking at two races we can determine whether voters use
natural experiment where two competing theories of voting
the same decision strategies across both races. Both contests
behavior can be tested.
were highly contested and high profile: $14.6 million was
spent by the mayoral candidates and $4.8 million was spent
expenditures approximately equal to that of the most expensive House
in the city attorney election.1 Clearly, spending was not
races; spending in the mayor’s race was greater than the most expensive
House races in 2002, but near that of the most expensive Senate races.
So by this measure, while the city attorney’s race was not as expensive
1 Campaign spending figures are from the Los Angeles City Ethics Com-
as the mayor’s race, the former still must be seen as a highly contested
mission. http://ethics.lacity.org/pdf/cf19892001.pdf. In comparison,
and high profile campaign.
according to data compiled by the Center for Responsive Politics
2 For comparative purposes, in the 1997 city attorney election, total
(http://www.opensecrets.org), in the 2002 election cycle the highest
spending was $2.7 million, and that was with an incumbent (Jim Hahn)
expenditure United States House election was the second district of West
in the election. While that amount is smaller than the 2001 figure, it still
Virginia, with $10.6 million spent, and the tenth-highest expenditure
is a considerable amount of money for a citywide office. In terms of
House race was the third district in Mississippi, with $4.5 million spent.
responsibilities, the city attorney holds a considerable amount of lever-
The Center for Responsive Politics reports that the most expensive 2002
age and influence in the city of Los Angeles. The Office of the City Attor-
Senate race occurred in North Carolina, with $26.9 million spent, and
ney is the largest municipal law office in the nation, with over 500 attor-
the tenth-most expensive Senate race was in Colorado, where $10.6
neys on their staff. The city attorney is responsible for prosecuting
million was spent. Thus, in these terms, the city attorneys race had
misdemeanors as well as providing counsel to the 35 departments and
agencies in the city of Los Angeles. In addition, the city attorney office
NOTE: The authors thank Jonathan Steinberg and George Waters for their
oversees a budget of approximately $96 million. For more information
input in earlier versions of this work.
on the city attorney position, refer to www.lacity.org/atty/atyoa1.htm.
Also, as just noted, Hahn was city attorney of Los Angeles for sixteen
Political Research Quarterly, Vol. 58, No. 2 (June 2005): pp. 203-218
years, from 1985 to 2001.

On one hand, theories of racially polarized voting sug-
whether ideology and issue positions even played a role in
gest that race is a primary determinant of vote choice in
these elections. Furthermore, in each of two elections on the
elections where a minority candidate is pitted against a
same day among the same sets of voters, a Latino candidate
white candidate. This theory is most relevant given the
competed against a white candidate.
undertones of racial fear that emerged in the last two weeks
The elections also provided variance in the election con-
of this election. James Hahn, the white mayoral candidate,
text. While both mayoral candidates were fairly liberal,
began broadcasting a television advertisement with an
Hahn was considered to be the more moderate of the two.6
image of a crack pipe held to a name followed by a grainy
The city attorney election, on the other hand, saw the can-
picture of the Latino mayoral candidate, Antonio Vil-
didacy of a moderate Latino, Rocky Delgadillo, while the
laraigosa.3 This image refers to a letter written by Vil-
white candidate, Mike Feuer, was more liberal (Marinucci,
laraigosa, requesting presidential clemency for a convicted
2001b).7 If race primarily determined vote choice, we
cocaine dealer. The advertisement then mentions that the
would expect that Latinos would vote for Villaraigosa and
cocaine dealer’s father had donated more than $6,000 to Vil-
Delgadillo, while whites would vote for Hahn and Feuer.
laraigosa’s campaign. While Hahn defended his advertise-
But, the spatial model predicts moderate and conservative
ment as factual and one that questions Villaraigosa’s trust-
Latinos and moderate and conservative whites voting for
worthiness and position on crime, both the media and
Hahn and Delgadillo, with liberal Latinos and liberal whites
Villaraigosa’s campaign argued that the advertisement
voting for Villaraigosa and Feuer.
played the race card, primarily appealing to racist whites in
Below we present evidence suggesting that the spatial
the San Fernando Valley.4 Mendelberg’s work (1997, 2001)
model has more power than previously believed in elections
would characterize this advertisement as an implicit racial
offering the opportunity for racial voting. Because of the
appeal, where this message would prime prejudiced whites
nature of this election, we use survey data to estimate voter
to vote in a racially prejudiced manner against the Latino
choice over pairs of candidates using bivariate probit analysis.
mayoral candidate. Thus, if voters were completely racially
Our results demonstrate that by observing two races involv-
polarized, we should observe all whites voting for the white
ing Latino candidates we nd that racial polarization is lower
candidate (Hahn) and all Latinos supporting Villaraigosa.5
than what we would infer had we studied a single race, and
On the other hand, the spatial theory of voting assumes
that the spatial model is relevant in white-Latino elections.
that voters rely on non-demographic cues: the correspon-
dence between the issue positions and ideology of the voter
and the candidates (Downs 1957). If there is no correspon-
dence between voters’ race or ethnicity and their issue posi-
tions, then the spatial model predicts that the observed rela-
Much of the literature on racially polarized voting has
tionship between votersrace and their vote choice is absent.
focused on the relationship between whites and blacks
However, in the case where issue positions and race are per-
(Gurin, Hatchett, and Jackson, 1989; Grofman and David-

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT