. 4
( 7)


ened, racial antagonism. According to the information model, the patterns
we see across cities should be a function of the quality of the information
that black leadership provides the white community. When experience
with a black mayor gives white residents clear, concrete information about
the effects of black leadership on the white community, changes in white
behavior should be pronounced. In contrast, when whites™ information
about black leadership is limited or uncertain, change in white behavior
should be minimal. But why do white residents get more information in
some cities than in other cities?
For whites, it is critical to know whether a black leader or some-
one else is responsible for the ongoing well-being of the white commu-
nity. The more that blacks are free to enact their own agendas and to

1 A factor that makes Dinkins™s failed reelection bid in New York particularly surprising is
the liberal nature of the white vote in that city. Why did an incumbent black Democrat
lose his reelection bid in a city that is generally so liberal? Part of the answer, I will argue,
is racial threat and fears of blacks eventually taking control. Dinkins was the recognized
leader of a large black community and that scared many whites “ even liberal whites.
When Dinkins sided with the African American community and failed to act decisively
against violent black incursions into the white community, it scared whites even more.
Who knew where the city was going? The result was a small shift in the white vote that
helped thwart Dinkins™s reelection bid. This is, however, not the entire story. An economic
recession, a terrorist bombing of the World Trade Center, and increased voter turnout in
more conservative Staten Island due to a referendum on that island™s secession all cut into
Dinkins™s liberal base.
Changing White Attitudes

overcome any constraints imposed by the white community, the more
whites will attribute their ongoing well-being to black leadership, and the
more insight they will gain into the interests of blacks and black leaders.
One implication of this logic, as I have explained, is that racial demograph-
ics play a critical role in whites™ perceptions of black leaders. Speci¬cally,
changes in white behavior should vary predictably across three types of
cities: majority black cities, majority white cities, and racially balanced
In cities where blacks are the majority of the population and the dom-
inant political group, it is easy for whites to link local conditions to the
interests and policies of black leadership. In these cities, when the world
under black leadership remains very similar to the world under white lead-
ership, many whites interpret this as a sign that they have little to fear from
black leaders. Blacks could have changed things and they chose not to. But
in other types of cities, the information provided by black incumbency is
more limited. Where whites are the clear majority and dominate the polit-
ical arena, the election of a black mayor does not represent a real transi-
tion of power from the white community to the black community. Whites
still do not know what would happen if the black community truly had
control. As a result, the actions of a black mayor in a majority white city
provide only limited information, and changes in white behavior are mod-
erate. In racially balanced cities, the information provided by black lead-
ership is even less clear. The election of a black mayor will not necessarily
lead to black control of the local political process, but a black mayoral
victory will likely put blacks on the cusp of power.2 This has two impor-
tant effects. First, it is likely to spur white elites into action. Rather than
cede control of local political resources, whites will mobilize to prevent
blacks from taking control. The result in many cases is policy paralysis.
Because no one is clearly in control and a black agenda is never realized,
most white voters receive little new information from the black mayoralty.
Second, the rough balance of power between the black and white com-
munities raises considerable uncertainty about subsequent elections. In
any future election, with the right outcome, a black-led coalition really
could gain control and take over the local political arena. With little new
information and the possibility of black control becoming much closer,
white opposition to black leadership is likely to increase in these cities.

2 In the racially balanced cities that are examined in this book, blacks held on average
one-third of the city council seats at the time the ¬rst black mayor was elected.
Learning Across Different Cities 95

testing the effect of racial demographics
on the white vote
To test these predictions, I divided the twenty-six cities with black chal-
lenger and black incumbent elections into three groups. Cities with pop-
ulations that were more than 55 percent white were coded as majority
white, those with populations that were 45 to 55 percent white were
coded as racially balanced, and those whose populations were less than
45 percent white were coded as minority white.3 I chose these cutoffs
because they adhere to previous empirical research.4 However, the exact
cutoff points chosen to distinguish each type of city are not critical.5
Table 4.1 illustrates that white reactions to black leadership are clearly
dependent on racial demographics in the ways predicted by the informa-
tion model. In minority white cities, white support for the same black
candidate increased by an average of 16.3 percent. This gain roughly
doubled the level of white support in these cities “ from 16.1 percent in
black challenger elections to 32.4 percent in black incumbent elections.
In majority white cities, white support increased by an average of only

3 The majority white cities are Charlotte, Cleveland, Kansas City, Los Angeles, Minneapolis,
Rockford, and Seattle. The racially balanced cities are Chicago, Dallas, Durham, Flint,
Houston, New Haven, New York, Philadelphia, and San Francisco. The minority white
cities are Atlantic City, Baltimore, Birmingham, Gary, Memphis, Newark, Oakland, and
4 Grofman and Handley (1989) have found that congressional districts that are 40 to 60%
black elect black representatives about half the time. Cameron and his coauthors (1996)
have suggested that the point of equal opportunity to elect a black candidate is generally
less than 50% black. The literature on schools has similarly found that at around 50%
white, the balance of power between white and black students is most precarious and the
intensity of competition between races is highest (Longshore 1988; Crain, Mahard, and
Narot 1982; Bullock 1976).
5 Alternate cutoffs (e.g. 40 to 60% white) lead to similar patterns of behavior across cities.
Also, whether one institutes cutoffs based on the size of the white population, the size of
the black population, or the relative size of the two populations, the results are essentially
the same. The main reason why these different speci¬cations lead to similar results is
that most of these cities had very small Asian American and Latino populations at the
time that they elected black mayors. On average, 92% of the residents of these cities
were black or white. Thus, focusing on the size of the white population is essentially the
mirror opposite of focusing on the black population. In large part because of the small
size of their populations, Asian American and Latino voters tend to play a limited role
in the elections under consideration. Of course, there are exceptions, like Chicago, where
Latinos can provide the margin of victory for either a white- or black-led coalition. Given
the relatively small size of the Latino and Asian American populations in these cities, it is
extremely dif¬cult to try to measure and assess their reactions to black leadership in any
systematic way.
Changing White Attitudes

table 4.1 Change in Political Behavior from Black Candidacy to Black
Incumbency: The Impact of the Racial Balance of Power

Change in:
Percent White Vote Margin of Percent Black Incumbent
for Black Candidate Victory Turnout Reelection Rate
Minority White 19.6 19.6 100% (8 of 8)
’7.5 ’7.7 ’4.4
Racially Balanced 63% (5 of 8)
Majority White 7.1 14.3 100% (7 of 7)

6.2 percent (from 42.2 percent to 48.4 percent).6 And in racially balanced
cities, white support actually decreased by an average of 2.9 percent.
Not surprisingly, these changes in the white vote affected the outcome
of each contest. Whether or not black incumbents won reelection was
greatly dependent on whether they faced a racially balanced, minority
white, or majority white electorate. In the minority and majority white
cities, black incumbents won reelection every time, while in racially bal-
anced cities, the black incumbent won reelection only about half of the
time (5 out of 8 cases). The margin of victory in black mayoral elec-
tions was similarly dependent on the racial balance of power. On average,
the margin of victory for the black candidate decreased by 7.7 percent-
age points in racially balanced cities. In contrast, the margin of victory
increased by an average of 19.6 percentage points in minority white cities
and 14.3 points in majority white cities. Measures of overall turnout con-
form to this same pattern. The turnout decline in majority or minority
white cities seems to suggest that the often intense white mobilization
against the black challenger had faded by the time the candidate came
up for reelection. The somewhat smaller turnout decline in racially bal-
anced cities, by contrast, suggests that white fears and concerns remained
relatively high in these cities even when blacks ran as incumbents.
It would seem, then, that white reactions to black leadership closely
mirror the predictions of the information model. In minority white cities,
whites became much more accepting of black leadership. In majority white

6 The more limited increase in white support in majority white cities might at least in
part be due to the fact that white support is slowly approaching a maximum threshold.
Candidates, after all, rarely get unanimous support. It is, however, worth noting that
mayoral elections are usually not close. The average margin of victory in mayoral elections
is 24 percentage points (Hajnal, Lewis, and Louch 2002). Given that successful mayoral
candidates typically win a large majority of the vote, it is unlikely that the 48% white
support we see here is pushing a maximum threshold.
Learning Across Different Cities 97

cities, change was positive, but not as strong. And, in racially balanced
cities, whites were just as opposed to black incumbents as they were to
black challengers. All of this suggests that there is a link between the
level of black control of the local political arena, information, and white

testing competing explanations
Could other more conventional political factors “ candidate quality,
endorsements, and the incumbent™s record “ account for the differences in
whites™ reactions in different kinds of cities? To try to ascertain if changes
in white behavior were in fact related to the information environment
rather than to other features of the local political arena, I undertook
more formal multivariate analysis of the change in the white vote. In this
analysis, I incorporated the same range of features of the local political
arena that I used in the investigation of black challenger and black incum-
bent elections in Chapter 2. To assess the impact of basic features of the
campaign, I took into account changes in newspaper endorsements, the
Democratic Party endorsement, and the quality of the white opponent.
To see if the black incumbent™s record affected the change in the white
vote, I included three features of the local environment that were likely
to have been most prominent in white voters™ minds, given the fears they
expressed when black challengers ¬rst entered of¬ce: change in per capita
income, change in median housing prices, and change in the local crime
rate. To determine if the policy actions of the black mayor also mattered,
I added a measure of the degree to which the black mayor™s administra-
tion shifted spending from development to redistribution “ presumably a
change that would most dramatically bene¬t the black community at the
expense of at least some of the white community.7 Finally, to investigate if
the racial content of the black candidate™s campaign in¬‚uenced the white
vote, I included a measure of the degree to which it shifted from being
race neutral to racially focused. Although I have tried to make this list as
comprehensive as possible, other factors in the local environment, such as
the local political culture, might also have played a role in these contests.
Details about the coding and sources of each of the variables are included
in the Appendix to Chapter 2.

7 In alternate tests, I also looked to see if greater pursuit of af¬rmative action in public hiring
affected changes in the white vote, but I found no link. The lack of an effect was likely due
to the fact that most black mayors pursued af¬rmative action at a very moderate pace.
Changing White Attitudes

The number of independent variables and the limited number of cases
once again stretch the limits of what an ordinary least-squares regression
can do, but the results do seem to ¬t a clear pattern: where blacks have
more ability to control local events, whites seem to learn more from their
experiences with black leadership and appear to be more willing to sup-
port black incumbents.8 After controlling for an array of variables related
to the campaign and the incumbent™s record, the results in Table 4.2 indi-
cate that racial demographics have a substantial effect on changes in the
white vote. All else being equal, the model suggests that white support
grows 12 percentage points more in minority white cities than it does in
racially balanced cities. The difference between white majority cities and
racially balanced cities is not nearly as stark, but there is at least some
indication that whites learn more in majority white cities than they do
in racially balanced cities.9 Table 4.2 also indicates that racial control is
not the only variable that in¬‚uences the white vote. Race enters into the
calculus of white voters in other ways. Black candidates who move away
from racialized campaigns toward more race-neutral campaigns tend to
garner a larger share of the white vote. Similarly, when fears of increasing
crime are not borne out by experience with black mayoral leadership,
white residents appear to grow more accepting of that leadership. None
of the other features of the incumbent™s record or the campaign is sig-
ni¬cantly related to the white vote, but it is worth noting that for every
variable in Table 4.2 the coef¬cient is in the right direction, and in some
cases, such as party and newspaper endorsements, is reasonably close to
being signi¬cant. It may be that better measures of the campaign and the
incumbent™s record would reveal clearer relationships.

racial demographics and the tone of mayoral elections
The extent to which racial demographics shape both the information envi-
ronment and white reactions can also be seen in the racial tone of elec-
tions. Changes in the tone or racial mood of campaigns in each type of
city follow a pattern very similar to the vote. In racially balanced cities,
the reelection bids of black incumbents tend to retain a high level of

8 Alternate tests dropping some of the independent variables and bivariate correlations led
to roughly the same set of ¬ndings as those demonstrated in Table 4.2.
9 In the model in Table 4.2, the coef¬cient for the white majority is positive but not quite
signi¬cant. Bivariate tests reveal a positive and signi¬cant correlation between change in
the white vote and a dummy variable for majority white cities.
Learning Across Different Cities 99

table 4.2 Determinants of Change in White Support for the Black

Change in White Support
for the Black Candidate
0.18 (0.07)——
White Minority Cityb
White Majority Cityb 0.07 (0.06)
Percent Black 0.00 (0.26)
Candidate Quality
’0.10 (0.11)
Change in Quality of White Opponent
Change in Democratic Party Endorsement 0.12 (0.14)
Change in Local Newspaper Endorsement 0.14 (0.11)
Racial Focus of the Campaign
’0.31 (0.13)——
Change in Black Candidate™s Racial Focus
Local Conditions
Change in Per Capita Income 0.49 (0.72)
Change in Median Housing Prices 0.18 (0.11)
’0.01 (0.00)—
Change in Crime Rate
’0.58 (0.45)
Change in Redistributive Spending
’0.24 (0.23)
Adj. R2 0.45
N 25
Note: OLS regression. Figures in parentheses are standard errors.
—— p < 0.01
— p < 0.05

p < 0.10
a All variables and their sources are described in detail in Appendix B.
b The excluded or comparison group is racially balanced cities (45% to 55% white).

White minority coded as <45% white. Majority white coded as >55% white.

racial tension. In Chicago, for example, when Harold Washington ran for
reelection in 1987, he faced opponents who did not hesitate to play the
race card. The racial theme of the slogan chosen by Washington™s primary
opponent, Jane Byrne “ “Jane Byrne: A Mayor for All Chicago” “ was
clear to all involved. Similarly, Ed Vrdolyak, Washington™s main oppo-
nent in the general election, made “bold use of racial themes throughout
his campaign” (Grimshaw 1992: 192). In Philadelphia, another racially
balanced city, there was also considerable racial hostility in the black
incumbent election. When Wilson Goode ran for reelection in 1987, the
Changing White Attitudes

local media saw the election as “a campaign in black and white” (quoted
in Adams 1994). Goode™s opponent, Frank Rizzo, was “pilloried by the
press as a race baiter” and at one point simply stated, “Vote white”
(P. Taylor 1987). Accounts in Cleveland, another racially balanced city,
also highlighted the role of race in the black incumbent™s ¬rst reelection
bid. In 1969, when Carl Stokes ran for reelection, the campaign focused
in large part on a police shootout in an African American neighborhood,
and, according to Levine, “despite disclaimers from both sides, the race
issue again dominated Cleveland politics” (1974: 59).
In contrast, in minority and majority black cities, black incumbent
reelection bids tended to be more subdued affairs that almost always
focused on the incumbent™s record and the traditional array of policy
issues that any candidate, black or white, would face. Norm Rice™s reelec-
tion in Seattle was dubbed by many as a foregone conclusion. Emmanuel
Cleaver™s reelection bid in Kansas City was greeted with widespread apa-
thy. And Richard Arrington™s reelection bid in Birmingham was so calm
and racially civil that the national press ignored it. Arrington himself
lamented: “Since we didn™t have any [racial] animosity, we didn™t get any
attention” (Russakof 1983). The tone of incumbent elections in these
cities suggests that uncertainty surrounding black leadership and fears
about what might happen after the next election had, at least in part,
been erased by the tenure of the ¬rst black mayor.
This pattern of change across cities has an important implication for
the literature on racial context and interracial dynamics. As discussed,
much of the scholarship on racial context has found that the larger the
local black population, the greater the level of racial animosity (see, e.g.,
Key 1949; Giles and Hertz 1994). And indeed, Chapter 3 demonstrated
a linear relationship between the size of the black population and the
negativity of white views. This chapter™s analysis suggests, however, that
once a black candidate assumes an important leadership position, the
simple linear relationship no longer holds.10 Because the reality of black
control turns out to be less menacing than the threat posed by the pos-
sibility of black control, racial tension appears to peak in cities where
black control is imminent but not yet in place. Thus, the racial divide
often remains exceptionally high in racially balanced cities and ironically
declines markedly in cities where blacks are more numerous and presum-
ably more able to take over.

10 For a similar argument about racial balance, see Blalock (1967) and Longshore (1988).
Learning Across Different Cities 101

table 4.3 The Impact of the Racial Balance of
Power on Black Incumbent Reelection Rates

City Demographics Incumbent Reelection Rate
Minority White 81% (61 of 75 cases)
Racially Balanced 65% (13 of 20 cases)
Majority White 77% (24 of 31 cases)

a broader range of cases
Information and racial control appear to play a key role in these twenty-
six cities, but can they account for electoral outcomes across a wide range
of cases? To answer this question, I once again looked at the outcome
of every black incumbent reelection bid in the twentieth century in every
city in the United States with over 50,000 people.11 When we shift to this
more complete set of cases, the same pattern emerges. Outcomes in black
incumbent elections across the United States have been strongly related
to a city™s racial balance of power. As Table 4.3 reveals, in minority white
cities, black incumbents win almost all of their reelection bids (81 per-
cent). In majority white cities, they are slightly less successful: they win
77 percent of their reelection bids. But in racially balanced cities, black
incumbents win only 65 percent of the time.
The data presented in this chapter suggest that when whites can hold
black leaders accountable for the lack of negative change that occurs
under their watch, then these leaders™ tenure in of¬ce will provide valuable
information to white voters about the effects of black leadership. This,
in turn, will lead to greatly increased white support and almost certain
reelection. But in racially balanced cities, black mayoral control is often
too precarious and racial competition too intense for white residents to
know who is to blame or credit for their continued well-being. In such
cities, white fears remain, elections continue to be highly racialized, and
black politicians have a more dif¬cult time getting reelected. In short,
information and uncertainty appear to play a critical role in the decisions
of white voters.

11 Each city is also coded as majority white, racially balanced, or minority white at the
time of the election. The cutoffs are the same as those employed for the smaller data
set: majority white (over 55% white), racially balanced (45 to 55% white), and minority
white (under 45% white). Alternate tests that break down cities based on the size of the
black population or the relative sizes of the black and white populations reveal similar
Changing White Attitudes

The pattern of white voting across cities that this chapter has iden-
ti¬ed may have practical implications for redistricting decisions. Many
have advocated the creation of more racially mixed districts, both because
blacks have a good chance of winning of¬ce in this type of district and
because they may maximize black substantive representation (Grofman
and Handley 1989; Lublin 1997; Cameron, Epstein, and Halloran 1996).
My results indicate, however, that black politicians may have a more dif¬-
cult time getting reelected in racially balanced arenas and that the creation
of racially competitive districts could actually lead to an increase in racial
tension.12 It is by no means clear that the pattern we see across cities
will be found in legislative districts or other types of geographic contexts,
and thus much more investigation of this phenomenon is required. Nev-
ertheless, those drawing the lines should at least begin to consider two
new concerns that have generally been previously overlooked: reelection
rates and the level of black-white con¬‚ict associated with different racial
demographics.13 Both of these concerns may lead legislators who favor
greater black representation away from supporting the creation of racially
balanced districts.
This chapter has shown that patterns of change in white support can be
linked to the amount of control that black incumbents are able to exercise
over the local political arena. The patterns evident here certainly ¬t the
expectations of the information model. Thus, there is little doubt that
the information model provides at least a plausible explanation of changes
in white behavior. But these simple correlations do not actually show us
how information is or is not translated by white residents in a given
community. In order to connect changes in white behavior more directly
to changes in the information environment, in the next two chapters I
examine two cases of black mayoral leadership in much more detail. The
goal of these two chapters is to follow the process of racial learning and
to directly link the actions of black incumbents to speci¬c changes in the
views and actions of white residents.

12 Interestingly, as noted earlier, studies at the school level also suggest that racial tensions
may peak in racially balanced communities (Longshore 1988).
13 These are obviously only two of the possible criteria by which a district can be judged.
One might also want to consider how district lines affect substantive representation,
descriptive representation, and a host of other issues (see Mansbridge 1999; Cameron,
Epstein, and Halloran 1996; Lublin 1997; Swain 1995).

Black Mayoral Leadership in Los Angeles

None of the data presented to this point allows us to follow the course
of black leadership in a single city to see daily, monthly, or even yearly
changes. This is an important omission, because learning under black
incumbents is a process. What we have seen so far is a glimpse into the out-
come of that process, but statistical data provide little insight into exactly
what information is provided by black leadership, how that information
is transmitted, and ultimately how it is interpreted by city residents. In
this chapter, I begin to address this omission by presenting more direct evi-
dence of the process of racial learning. Speci¬cally, I follow the actions of
black mayoral leadership and the attitudes and actions of white residents
before, during, and after the transition from white to black leadership
in Los Angeles. The goal is to demonstrate as clearly as possible how
information from black incumbency is translated into changes in white
attitudes and behavior.
I focus on Los Angeles for both practical and theoretical reasons. On
a practical level, data constraints con¬ned my choices to cities that had
substantial empirical records of white views during a black mayoralty.
Unlike almost all other cities that have experienced black mayoral repre-
sentation, Los Angeles has mayoral exit polls that allow for analysis of
the white vote in several contests, and several of these polls included
questions assessing white racial concerns. This allows me to gauge
the effects of racial views on the white vote at different points in time.
Coupled with rich primary and secondary accounts of Mayor Tom
Bradley™s record in Los Angeles, these polls allow for perhaps the most
comprehensive assessment of white reactions to black leadership avail-
able in any city.
Changing White Attitudes

On a theoretical level, moreover, Los Angeles under Tom Bradley offers
an interesting test case. When he was elected in 1973, Bradley became
one of the ¬rst black mayors in the country, so his tenure marked one
of the ¬rst times that white Americans could gauge the consequences of
black leadership in an important elected of¬ce. Thus, we might expect
learning to be particularly pronounced here. In addition, Bradley™s long
tenure as mayor (twenty years) allows for a more re¬ned assessment of
how white attitudes and actions change over time. Even if the process of
learning is a slow one, it should still be evident in Los Angeles. A more
extensive account of the motivations behind my case selection is included
in Appendix C.
In this chapter, I assess three different aspects of the learning process.
First, to what extent did a lack of information about the consequences of
black leadership play a role in black challenger elections? If the informa-
tion model is accurate, a lack of concrete information about the conse-
quences of black leadership for the white community should be one of the
central features of black challenger campaigns. To test this hypothesis, I
detail the views and concerns of white residents during Bradley™s two cam-
paigns to become the city™s ¬rst black mayor, and evaluate the role racial
concerns played in the white vote in these two contests. Second, what
information did black incumbency provide? Here, I review Bradley™s pol-
icy initiatives and their impact on the economic vitality of the white com-
munity to show that Bradley™s tenure did little to harm white Angelenos.
And third, was information provided by black leadership translated into
changes in white attitudes and behavior? In particular, did white resi-
dents update their beliefs about blacks and black leadership? Again, if the
information model is accurate, at least some whites should have taken the
limited impact of black leadership on the white community as a sign that
black leaders were not out to get them, racial concerns should have played
a declining role in the white vote over time, and whites should have been
more willing to vote for a black candidate in subsequent elections. I test
this last set of hypotheses by looking at changes over time in overall white
support for Bradley, in the determinants of the white vote, and in the sen-
timents expressed about Bradley. Once again, I contrast the information
model with a racial prejudice model and a white backlash model. If the
racial prejudice model ¬ts Los Angeles, there should have been no change
in white views or actions under a black mayor. If the backlash model is
accurate, whites should have increased their efforts to try to reverse black
gains and oust black leadership.
Black Mayoral Leadership in Los Angeles 105

For this narrative of Los Angeles politics, I rely primarily on newspaper
reporting and primary accounts of events in the city. However, I also
draw on previous analyses when they provide further insight. All of this
is supplemented with in-depth analysis of two public opinion polls: one
conducted in 1969, the year of Bradley™s ¬rst mayoral bid and one from
1980, the midway point of Bradley™s mayoral tenure. Both are described
in more detail in Appendix C.

black challenger elections: limited information
and white fear
To begin to assess the information model, I examined Tom Bradley™s two
bids to become the ¬rst black mayor of Los Angeles (his unsuccessful bid
for the mayoralty in 1969 and his successful campaign in 1973) to see if
uncertainty about black leadership led to racial fears and opposition to
Bradley™s candidacy. The results closely ¬t the information model. Almost
every aspect of both elections points to acute racial concerns about Tom
Bradley and widespread fear about how black leadership could affect the
white community.
While the evidence is anecdotal in nature, the white Angelenos who
were interviewed in the days and weeks before the 1969 election were
often very clear about their fears of black leadership. These individu-
als regularly expressed deep-seated concerns about the potential conse-
quences of black leadership for the white community. The comments of Jo
Ann Des Ruisseaux were typical: “I just don™t like all these Black Panther
people that are hanging around [Bradley™s] campaign. I know he denies
it, but where there™s smoke, there™s ¬re” (Reich 1973b). Many focused
on black in-migration as a real threat. One white resident lamented:
“If we have a colored mayor we™ll have colored people pushing us
out of the city. The whole city will be black if Bradley wins” (Reich
1973a). A citizen from Encino echoed this concern about a black victory:
“You know what this means? There™ll be blacks all over this place next
month. They™ll be all over Encino. You™re not gonna recognize this town”
(Seidenbaum 1973). Others did not mention speci¬c threats, but sug-
gested that the consequences of Bradley™s election could be dire for the
white community.
Perhaps the most persuasive evidence of the important role that uncer-
tainty and white racial concerns played in these two contests comes from
analysis of the white vote itself. Whether whites voted for or against
Changing White Attitudes

table 5.1 Key Factors Driving the Vote for the Challenger Tom Bradley1

Change in the Probability
of Voting for Bradley (%)
Bradley Favors Blacks 67
Concerned About Black Gains “ Not Concerned
Satis¬ed With Public Services “ Dissatis¬ed Not Signi¬cant
Satis¬ed With Economic Gains “ Dissatis¬ed Not Signi¬cant
Conservative “ Liberal 27
Republican “ Democrat 37
Racially Prejudiced “ Not Racially Prejudiced Not Signi¬cant
1 Predicted probabilities derived from logistic regression in Table C.1 in Appendix C.

Bradley depended more than anything else on their concerns about
Bradley and black leadership. One poll from the 1969 runoff between
Bradley and Sam Yorty “ who was running for reelection “ is particularly
illuminating. In Table 5.1, I present an account of the main factors driving
the white vote in that election.1 The predicted probabilities in the table
are derived from a logistic regression that assesses the effects of racial
concerns, Yorty™s record as an incumbent, racial prejudice, political ideol-
ogy, and socioeconomic status. The survey, variables, and full model are
detailed in Appendix C.2
Several important conclusions emerge from this analysis. First, con-
cerns about how black leadership would affect the white community dom-
inated white voters™ decision making. Whether or not whites supported
Bradley depended principally on the simple question of whether or not
they thought Bradley would serve black interests at the expense of white
interests. If a white respondent thought that Bradley “would show more
favoritism to his supporters than most other mayors,” that person was
67 percent less likely to support Bradley than if the voter thought Bradley
would be evenhanded. Similarly, if whites were concerned about black
political and economic gains and felt that blacks would “push themselves

1 There is, unfortunately, no comparable survey of white voters in 1973.
2 The survey is a pre-election poll of white suburbanites in Los Angeles conducted by the
National Opinion Research Center shortly before the run-off election between Bradley
and Yorty.
Black Mayoral Leadership in Los Angeles 107

where they™re not wanted,” they were 54 percent less likely to vote for
Bradley. Fears about losing out to blacks clearly were central to the white
The other important and unique aspect of the Bradley-Yorty contest
was the fact that the white vote was unrelated to Yorty™s record as an
incumbent in the mayor™s of¬ce. As Table 5.1 indicates, whether or not
white residents were satis¬ed with city services under Yorty and whether
or not they were satis¬ed with their personal economic gains over the
previous ¬ve years had little effect on whether white voters chose Bradley
or Yorty. The white vote was not about the white candidate or the past.
It was instead about the black candidate, the future, and concerns that
black leadership might harm the white community.
At the same time, Table 5.1 reveals that racial concerns were not the
only factor that mattered in the white vote. Nonracial concerns, including
political ideology and party identi¬cation, played a modest role in the
contest. Liberals and Democrats were somewhat more likely than con-
servatives or Republicans to support Bradley “ although the difference
was not as large as one would traditionally expect. Self-identi¬ed liber-
als were only 27 percent more likely to support Bradley than those who
viewed themselves as conservative. In the end, Bradley received consider-
ably more white support than other black challengers who were gaining
of¬ce in other big cities at around the same time. Overall, he garnered
32 percent of the white vote in 1969 and 46 percent in 1973 (Halley,
Acock, and Greene 1976). By contrast, successful black challengers in
Cleveland, Gary, Newark, and Atlanta, the four other big cities electing
black mayors around the same time, won only between 15 and 17 percent
of the white vote.
The fact that nonracial concerns did sway some white voters in Los
Angeles and the relatively widespread white support for Bradley when
he ran as a challenger is exactly what one would expect given that the
black community in Los Angeles was much smaller and posed less of a
threat than the black population in these other majority black cities. In
1969, Los Angeles was a majority white city, and African Americans made
up only 17 percent of the population. Los Angeles™ demographics meant
that whites would likely retain ¬rm control of the local political arena,

3 Equally importantly, as Table 5.1 shows, being prejudiced against blacks had no signi¬cant
effect on the vote. Speci¬cally, holding negative stereotypes of blacks (thinking blacks
are less intelligent than whites) played no direct role. It was not simple prejudice that
structured the vote, but expectations about the future and fears about black leadership.
Changing White Attitudes

even if Bradley were elected. Uncertainty played a role in Los Angeles
but concerns about black leadership in this majority white city were less
pronounced than were white fears in cities where blacks really could “take
The media in Los Angeles echoed this analysis of the white vote. Most
of the stories run by the Los Angeles Times highlighted in one way or
another the important role that white fears about black leadership played
in Bradley™s bids to become mayor.4 Kenneth Reich, in particular, saw “a
distinct note of concern about a black man in the mayor™s of¬ce” (1973a).
His interviews of white Angelenos “indicated that the racial issue remains
highly important this year.”5
White concerns about what might happen if blacks were allowed to
take over also came through in the campaigns that Bradley and his oppo-
nent ran. Yorty, in particular, designed his two campaigns against Bradley
to highlight all of the unknowns surrounding a black victory. In both 1969
and 1973, the theme of Yorty™s campaign was typi¬ed by the following
statement: “We know what kind of a city we™ve got. We don™t know what
we might get [if Bradley is elected]. So we™d be taking quite a chance
with this particular kind of candidate” (Bergholz 1973b). Time and time
again, Yorty accused Bradley of being a radical who was anti-police and
pro-communist. He repeatedly raised questions about the future of Los
Angeles under black leadership. “Will your family be safe?” asked one
advertisement. “Will your city be safe with this man?” queried another
(Bollens and Geyer 1973). There was, Yorty argued, a real chance of
“losing the city” (Boyarsky 1973c). In short, Yorty clearly felt that the
best way to defeat his black opponent was to play on white fears about
the unknowns surrounding black leadership.
Yorty was not the only candidate who realized that white Angelenos
were afraid of black leadership. Bradley also knew that concerns about
black leadership were widespread, and he did everything he could to try
to assuage those fears in his campaigns as a challenger. By avoiding any

4 See e.g. Boyarsky 1973a; 1973b; 1973c; 1973d; 1973e; Bergholz 1973a; Paegel 1973; and
Reich (1973a; 1973b.
5 More extended scholarly accounts that were written in the years after the elections reaf-
¬rmed the central role of racial fears. In fact, it is quite remarkable that of the dozen or
so studies that examined the two elections in depth, every single one concluded that fear
and uncertainty plagued Bradley™s bid for the mayoralty. The following accounts all point
to the critical role uncertainty played in Bradley™s two challenger elections: Bollens and
Geyer 1973; Hahn, Klingman, and Pachon 1976; Halley, Acock, and Greene 1976; Jeffries
and Ransford 1972; Kaufman 1998; Kinder and Sears 1981; Litwin 1981; Maullin 1971;
Payne and Ratzan 1986; Pettigrew 1972; Robinson 1976; Sears and Kinder 1971; Shiesl
1990; Sonenshein 1989; 1993; Watson 1984.
Black Mayoral Leadership in Los Angeles 109

mention of racial issues, downplaying his connections to black activists,
and focusing his campaign on law and order, Bradley made it clear that his
was not a “black” campaign. He did not bring in any well-known black
leaders from outside the city, he campaigned largely in white neighbor-
hoods, and he did not talk about the issue of race except to occasionally
deny that it was an issue. As Sonenshein put it, “Bradley forces went to
great lengths to reassure white and Hispanic voters that Bradley would
be fair to all” (1989: 343).
After he lost in 1969, Bradley and his advisors were certain that
racial fears had cost him the election. They were determined not to
let white fears dominate his second challenger bid (Payne and Ratzan
1986b). Thus, in 1973, Bradley™s efforts to reassure white voters greatly
increased. He repeatedly stressed his ties to the Los Angeles Police Depart-
ment and focused much of his campaign on conservative law and order
issues (Boyarsky 1973c). When Yorty focused media attention on a Black
Panther who had endorsed Bradley, Bradley™s response was swift and
decisive: “I do not seek, I do not want, and I reject the endorsement.
I have always been opposed to such a philosophy” (Paegel 1973).
What all of this evidence, both quantitative and qualitative, suggests is
that uncertainty about black leadership was widespread among white res-
idents in Los Angeles when Bradley ran as a challenger. Bradley may have
had a record as a moderate on the city council, but whites in Los Angeles
had never experienced an African American in control. The experiences
that white Angelenos did have “ the Watts Riot, periodic violent racial
¬‚are-ups, and growing demands from the black community “ undoubtedly
led many to expect the worst.

the information effects of bradley™s tenure
In this section, I review Bradley™s tenure in an effort to see what his agenda
looked like, how well he was able to enact that agenda, and how the
economic and social well-being of white Los Angeles changed under his
mayoralty. The bottom line is that if whites were watching Bradley to
see what black leadership meant to the white community, what they saw
and experienced under Bradley suggested that black leadership was not a
serious threat to their well-being.

Bradley™s Pro-Development Agenda
Although there were some areas in which Bradley pushed for modest gains
for blacks, the main theme of his administration, according to almost
Changing White Attitudes

all observers of Los Angeles politics, was economic development, and
in particular downtown development (Saltzstein 1986; Sonenshein 1993;
Regalado 1991; 1992; M. Davis 1992). As part of this pro-development
agenda, Bradley maintained a ¬scally sound administration. In his ¬rst
years, he cut city expenditures by 10 percent and was able to balance the
city budget without raising taxes (Payne and Ratzan 1986b). There was
also no real redistribution of income or any real change in spending pri-
orities under Bradley (Regalado 1992; Jackson and Preston 1994; Ander-
son 1996).6 Over the twenty-year period of his tenure, local government
redistributive expenditures grew only 0.9 percent as a percentage of total
government expenditures (Bureau of the Census Bureau 1964“2003).
The core of Bradley™s policy agenda had little, if anything, to do with
race. Bradley, in fact, tried hard to avoid racial controversies and regularly
vowed to serve all racial groups. In his ¬rst speech as mayor, he pledged
that he would “be mayor of all of Los Angeles” (Payne and Ratzan 1986a:
135). And on most racial issues, Bradley™s regime was silent (Jackson 1990;
Sonenshein 1993). As Sonenshein has noted, “On such tinderbox issues
as school busing, Bradley has been utterly invisible” (1990: 40). In 1979,
when an LAPD of¬cer shot Eulia Love, a black woman, aggravating racial
tensions, Bradley™s response was “feeble,” according to many observers
(Anderson 1996). Bradley™s administration did provide some real, con-
crete bene¬ts for blacks in the city. In a regime that many felt was at
least “initially transformative,” Bradley fought to ensure greater over-
sight over the Los Angeles Police Department, and he attempted to open
the doors of city government to previously excluded groups (Sonenshein
1993). Minority contracting expanded, and more blacks were hired as
commissioners and in other positions of authority (Jackson and Preston
1994).7 Bradley also helped to increase the number and size of federal
grants for poverty and public housing spending when these funds were
widely available in the 1970s (Keiser 1997; Sonenshein 1993). But these
efforts were modest. Under Bradley, black employment in city govern-
ment only increased from 20 percent in the Yorty years to 24 percent in
the 1990s (Sonenshein 1993; Joyce 1994). Indeed, it is unclear whether
af¬rmative action in Los Angeles proceeded at a faster pace than it did

6 The biggest change in government under Bradley, according to Sonenshein (1993) was the
incorporation of business. As one Chamber of Commerce executive put it, “Business was
made a managing partner in running the city” (Litwin 1981: 87).
7 Black commission appointments increased from 6% under Yorty to 20% in 1991
(Sonenshein 1993).
Black Mayoral Leadership in Los Angeles 111

in other cities without black mayors (Jackson and Preston 1994; Litwin
1981; Eisinger 1983).
In short, Bradley did not show any sign that he was interested in a
“black takeover.” All of these actions “ and inactions “ on the part of the
Bradley administration likely served to reassure white Los Angeles that
black leadership was relatively safe for the white community.

Ongoing White Prosperity
More importantly, the end result of Bradley™s tenure was ongoing white
prosperity and no change in the relative economic status of white and
black Los Angeles. If anything, whites gained under Bradley, while much
of the black community fell behind. Los Angeles experienced an extended
period of economic expansion during Bradley™s tenure (Litwin 1981;
Oliver, Johnson, and Farel 1993; Ong and Blumenberg 1996; Regalado
1992; Schwada 1989; Sojo and Scott 1996). At a time when many other
cities went downhill, Bradley helped to transform Los Angeles into the
largest industrial center in the United States (Sonenshein 1993). Over-
all, the median family income of the city™s residents grew by 5.3 percent
between 1969 and 1989, marginally better than the average for urban
America (Ong and Blumenberg 1996). Importantly for the white popula-
tion, these gains had a disproportionate impact on the upper end of the
wage distribution. Of the hundreds of thousands of new jobs created dur-
ing Bradley™s administration, most were for highly skilled, well-educated
white workers (Ong and Blumenberg 1996). Most of the development
occurred downtown, where white business interests bene¬ted (Sonenshein
1993). On the other end of the spectrum, in Los Angeles™ poor black com-
munities, there were limited positive developments; if anything, economic
fortunes deteriorated under Bradley.8

8 Bradley™s mayoralty did little to alleviate poverty and the variety of social ills that
plagued much of black Los Angeles (Anderson 1996; Bunch 1990; M. Davis 1992;
Keiser 1997; Jackson and Preston 1994; Oliver, Johnson, and Farrel 1993; Ong and
Blumenberg 1996). Areas such as Southcentral Los Angeles suffered a slow decline. In
Southcentral between 1973 and 1990, the poverty rate, the unemployment rate, and
the proportion of residents on welfare all grew (Ong and Blumenberg 1996). By 1990
in Southcentral, unemployment exceeded 50%, welfare dependency stood at 25%, the
poverty rate was 30%, and 56% of the adult population were high school dropouts (Los
Angeles Times 1992). In other words, the efforts of Bradley™s administration “stopped
short of the population most in need” (Anderson 1996: 351). The fact that Bradley had
not met black expectations was con¬rmed by considerable disenchantment within the
black community, which became more and more evident as Bradley™s tenure went on
(M. Davis 1992; Sonenshein 1990).
Changing White Attitudes

The racial status quo that had seemed to many to be threatened by
the onset of black leadership survived the twenty years under Bradley.
A variety of economic and social indicators revealed no change in the
relative well-being of blacks and whites during Bradley™s tenure. In 1969
in Los Angeles, black male median earnings (controlling for education
and labor force experience) were 70 percent of white earnings. In 1989,
the ¬gure was a virtually identical 69 percent (Ong and Blumenberg
1996). Income inequality data reveal an increasing gap between wealthy,
mostly white Los Angeles and poor, disproportionately black Los An-
geles over the same period. Between 1969 and 1989, the GINI coef¬cient
grew from 0.37 to 0.44, and the poverty rate increased from 10.9 to 15.1
percent (Ong and Blumenberg 1996). As the Los Angeles Times noted in
1989, “An already yawning racial gulf between rich and poor is growing”
(2: 6). For anyone who watched, the lesson was clear: whites could sup-
port blacks without hurting their own well-being.9

did whites learn in los angeles?
Did whites, in the end, learn this lesson, updating their beliefs about
Bradley and black leadership? If the information model is correct and
Bradley™s tenure did demonstrate to white residents that black leadership
was not in con¬‚ict with white interests, then uncertainty about Bradley
and black leadership should have declined. And as uncertainty declined,
perceptions about the meaning of black leadership should have changed,
racial concerns should have faded in importance in subsequent elections,
and more and more whites should have supported Bradley. All of this is
borne out by the evidence.

Growing White Support and Falling Turnout
The most obvious change in white behavior was increased white support
for Bradley™s leadership. When Bradley was ¬rst elected, he faced record
white turnout, and most white voters voted against him. But as Table 5.2

9 It is also important to note that Bradley™s governing coalition faced little serious opposi-
tion during most of his tenure (Jackson 1990; Sonenshein 1993). Even though Bradley™s
mayoral powers were somewhat limited, what he proposed was often pretty close to
what Los Angeles got. Thus, white residents would have had a hard time pointing to
anyone else in the city as being responsible for the direction of policy under Bradley™s
Black Mayoral Leadership in Los Angeles 113

table 5.2 White Support for Bradley Over Time

White Vote for
Bradley (%) Voter Turnout (%)
1969 32 76
1973 44 75
1977 53 42
1981 58 45
1985 62 35
1989 47 23

illustrates, as white residents gained more and more experience with black
leadership, most of this opposition faded. In Bradley™s ¬rst reelection bid,
white support increased to 53 percent, and he won all but one city council
district. By 1979, polls showed Bradley with a 76 percent approval rat-
ing. One year later, the Los Angeles Times exclaimed, “Bradley is, within
the liberal community, unassailable” (Sonenshein 1993: 179). In 1981,
Bradley™s approval reached 85 percent, and he won 58 percent of the
white vote that year. Four years later, he garnered a record 68 percent of
the city vote, capturing 62 percent of the white vote and winning every
district “ including the conservative white Valley districts that had mobi-
lized in 1969 and 1973 in such great numbers to prevent what they thought
was a black takeover. It was only after scandals and a poor showing in
the California gubernatorial race tarnished his reputation that Bradley™s
winning percentage dropped for the ¬rst time, in 1989. He chose not to
run for a sixth term in 1993.
Remarkably, over a sixteen-year period, Bradley™s support in the white
community nearly doubled. A losing black challenger had been trans-
formed into a ¬ve-term mayoral juggernaut.10 By the end of his tenure,
Bradley™s support among white residents was not too different from his
support among black Angelenos. In 1990, blacks on average rated Bradley
as a 68 on a 100-point feeling thermometer. Whites weren™t that far behind

10 Analysis of the mayoral vote in key city council districts con¬rms Bradley™s grow-
ing support among all sectors of white Los Angeles. Over the ¬ve elections from
1969 to 1985, Bradley™s support among white liberals grew from 51 to 58 to 62 to
69 to 74%. Among conservatives, the increase in support was not nearly as dra-
matic, but it was nevertheless impressive. Over the same ¬ve elections, conserva-
tive white support for Bradley grew from 31% to 50%. Among moderate white and
Latino wards, Bradley™s support went from 44% in 1973 to 60% in 1985 (Sonenshein
Changing White Attitudes

at 57.11 In short, by the end of Bradley™s tenure, something resembling
racial consensus had crept into the mayoral politics of the city.
The direction of the vote was not the only sign in Los Angeles that racial
concerns surrounding black leadership had declined. Another possible
indication of diminished fear was the steep drop in voter turnout. Turnout
in Bradley™s mayoral contests plummeted from a record high of 75 percent
in 1969 to 42 percent in 1977, 37 percent in 1981, 35 percent in 1985, and
23 percent in 1989. The 23 percent turnout in 1989 represents a record
low for the city, and it occurred despite the fact that many observers
felt that Bradley had a real chance of losing that year. The fear that had
motivated so many white residents to try to prevent a black takeover
appears largely to have disappeared over the course of Bradley™s tenure.

The Declining Role of Race Among White Voters?
A less obvious but equally important change under Bradley was the dimin-
ishing role racial concerns played in campaigns and in the white vote. Just
as the information model predicts, the racial concerns that dominated the
calculus of white voters in Bradley™s two challenger candidacies slowly
faded with time. As each year went by and whites gained more expe-
rience with black leadership, Bradley™s record as an incumbent became
more and more critical to white voters.
Bradley™s ¬rst reelection bid in 1977 began the trend and is particularly
illustrative. In many ways, Bradley faced a dif¬cult electoral context that
year. In the interim four years, Los Angeles voters had moved markedly
to the right. As the election results would later show, in most of the city™s
municipal and school board elections, voters opted for more conservative
candidates than they had in the past (Sonenshein 1993). Moreover, dur-
ing the campaign, Los Angeles was facing a racially divisive school busing
controversy, a highly unfavorable setting for any black candidate.12 Alan
Robbins, Bradley™s chief opponent, tried to use this context to his advan-
tage. Just as Yorty had done before him, Robbins attempted to make race
relevant in the campaign, repeatedly trying to incite white racial concerns
about Bradley and black leadership by raising the issues of crime and

11 The increase in white support for Bradley is even more impressive when one considers the
conservative drift of the Los Angeles electorate. Bradley was increasingly successful dur-
ing a period in which Los Angeles™ white voters favored ¬scally conservative propositions
(e.g. Prop. 13 in 1978, Prop. A in 1981, and Prop. 1 in 1985).
12 All through the campaign, the busing controversy received more coverage than the cam-
paign did.
Black Mayoral Leadership in Los Angeles 115

busing, among others. He accused Bradley of allowing the school system
“to be set on an almost irreversible path toward extensive forced bus-
ing” (Reich 1977a). He criticized Bradley™s cuts in police spending, and
he accused Bradley of being soft on crime. He even went so far as to hand
out thousands of rape whistles. As one city hall observer noted, “Bradley
is black and the whole Robbins campaign is subtly directed toward play-
ing upon it” (Reich, 1977a). In the end, though, Robbins™s efforts failed.13
Although the attention that busing received indicated that race was still
an important issue in Los Angeles, the results of the election suggested
that most whites did not fear Bradley. With Bradley running as an incum-
bent, there was no run-off. There was no widespread white mobilization.
In short, there was no contest. White voters supported Bradley in large
numbers, white business provided Bradley with strong ¬nancial backing,
and turnout declined precipitously. In contrast to the 1969 and 1973 elec-
tions, racialized voting did not follow a racialized campaign.
On this point, most observers agreed:

There were two overriding reasons for Mayor Tom Bradley™s overwhelming
victory “ his own substantial record of achievement over the past four years and
the diminution of race as a factor in Los Angeles politics. The fact that Bradley is
black was signi¬cant in his two previous campaigns for mayor. There were simi-
lar, more indirect efforts this year to place a negative emphasis on race. . . . But the
voters, to their credit, saw more merit in Bradley™s pledge to work for peaceful
compliance. . . . We also commend the voters of this city for appraising Bradley on
the basis of performance, not of race. (Los Angeles Times 1977)

Even Bradley agreed that racial concerns had largely dropped out of the
equation. In a Los Angeles Times article entitled “Race Banished as Vote
Issue,” Bradley was quoted as saying, “You know that this city could
rise above race, could rise above economic circumstances, could rise above
politics, to elect a mayor on his quali¬cations, character, and con¬dence
in the progress of ideas” (Reich 1977b). Bradley went on to claim that
white voters had lost their fear of black leadership. They had learned that
a black mayor was not the end of the world: “This time, there was not
such an important issue to be proved. . . . [White voters] were sure enough,
ready, and willing to vote for a man who was black” (Reich 1977b). In
short, the 1977 election provided ample evidence that experience under

13 This time, against Robbins, Bradley did not even have to counter with a conservative
law and order campaign, as he had done before, against Yorty, to assuage white fears.
Instead, he simply campaigned on his record in of¬ce.
Changing White Attitudes

Bradley as mayor had allayed white fears concerning Bradley the black
Subsequent elections told a similar story. The 1981 election was an
almost exact repeat of 1977. In 1981, Bradley faced his old nemesis,
Yorty. And Yorty went with the same white fear campaign he had used
in 1969 and 1973. Race, crime, and fear were the order of the day for
the campaign. “People are afraid in this town. They™re afraid to go out at
night and afraid to go out in the daytime,” Yorty declared (Dowie 1981).
Though the campaign was the same, the reaction was not. This time
around, Yorty was sharply criticized for his racist attacks and his efforts
to incite white fear. As the campaign went on, in fact, Yorty was forced
to tone down his attacks (Payne and Ratzan 1986b). Also in contrast to
1973, Bradley did not respond to Yorty™s vitriol. Bradley™s commercials
did not mention Yorty or his tactics. There was none of the law and order
campaign that Bradley had used in 1973 to assuage white fears. There was
also little mention of his 21-year career on the LAPD. Bradley ran on his
record as mayor, and the results were impressive. He won 64 percent of
the vote to Yorty™s 32 percent, and he carried every council district in the
city. Bradley became the ¬rst mayor of Los Angeles to win a third term
without a run-off. A Los Angeles Times editorial applauded the city™s
white voters for ignoring Bradley™s race: “Just as encouraging was the
absence of race as a dominant issue in Sam Yorty™s battle to unseat the
black mayor” (Los Angeles Times 1981).
The story of the 1985 and 1989 mayoral elections is somewhat dif-
ferent, but the conclusion is the same. In both elections, race and racial
issues took a back seat to other city and voter concerns. In 1985, for the
¬rst time, Bradley™s main opponent did not use race in his campaign. John
Ferraro did run a negative attack campaign, but the issue was never race.
Unlike previous elections, there was a plethora of issues to choose from.
Ferraro tried to win votes by coming out against a city metro line and by
voicing opposition to a proposed property tax hike. The two big issues in
the election turned out to be Bradley™s support for offshore oil drilling and
whether or not Bradley would run for governor the next year. In Payne
and Ratzan™s view, “The issues were the focus of the campaign, and race
seemed, for the ¬rst time, to be of little concern to either the candidates or
the voters” (1986b: 329). In 1989, the names of the challengers changed,
but the script was largely repeated. Race had ceased to be the central
element in Los Angeles mayoral elections.
Empirical analysis of local newspaper coverage helps to con¬rm the
declining role of race in mayoral campaigns. In Bradley™s challenger
Black Mayoral Leadership in Los Angeles 117

table 5.3 Main Factors Driving Approval of Bradley

Impact on Support for:
Bradley the Bradley the
Challenger Incumbent
Concerned about Black Gains “ Not Concerned 58% Not Signi¬cant
Not Signi¬cant ’46%
Satis¬ed with the City Conditions “ Dissatis¬ed
Note: Predicted probabilities derived from logistic regressions in Table C.2 in Appendix C.

campaigns, race was the issue. Both in 1969 and 1973, race dominated
campaign coverage (Graber 1984). Far outweighing all other issues, race
was mentioned in half of all coverage of the 1973 election (Sylvie 1995).
For anyone who cared to listen, read, or watch, Bradley™s mayoral chal-
lenge was more than anything else a racial battle between a black and a
white candidate. But all of this changed after Bradley was elected mayor.
In Bradley™s mayoral reelection bids, the media slowly ceased to focus
on race. Graber™s (1984) analysis of campaign coverage in Los Angeles
reveals a substantial decrease in the frequency with which the issue of
race was raised in local newspapers. Bradley™s race had been replaced by
Bradley™s record.

views of bradley as challenger and incumbent
Another way to see if experience with black leadership fundamentally
altered white views in Los Angeles is to compare the factors driving white
support of Bradley as a challenger with the factors driving white support
of Bradley as an incumbent. If the information model is accurate, this
kind of comparison should reveal a sharp decline in the importance of
racial concerns and a sharp increase in the importance of the incumbent™s
record. This is exactly what we see in Table 5.3. The table contrasts the
factors driving white approval of Bradley in 1969, when he ¬rst ran as
a challenger, with the factors driving white approval of Bradley in 1980,
after he had already led the city for seven years.14 The predicted proba-
bilities in the table are derived from two regressions that assess the effects
of racial concerns, the incumbent™s record, racial prejudice, political

14 I focus on the 1980 poll because it is the only citywide poll during Bradley™s administration
that includes measures of white racial attitudes.
Changing White Attitudes

ideology, and individual socioeconomic status. The surveys, variables,
and full models are detailed in Appendix C.
The contrast between white views and concerns before Bradley was
elected and white views and concerns seven years into Bradley™s tenure is
dramatic. As the ¬rst column of Table 5.3 illustrates, when Bradley ran as
a challenger, the election was largely about racial concerns. Fears about
the possible consequences of black leadership dominated the white vote,
and the incumbent™s record played no role. Whites™ support for Bradley,
the black challenger, over Yorty, the white incumbent, was unaffected
by their feelings about how well the city was doing. Rather, support for
Bradley was almost wholly shaped by concerns about the future and the
fear that Bradley would serve black interests.
In sharp contrast, views of Bradley the incumbent were not signi¬cantly
tied to racial concerns. The racial fear that had shaped the white vote in
Bradley™s challenger election appears to have faded as whites gained expe-
rience under Tom Bradley and as the reality of black mayoral leadership
proved better than expected. Even more important, as the information
model would predict, support for Bradley as an incumbent was largely
a function of the information that his tenure had provided. If whites
thought the city had fared well under Bradley, they generally supported
him. Those who felt “things in Los Angeles” were going well were 46
percent more likely to have a favorable view of Bradley than those who
felt that Los Angeles was faring poorly. In other words, the key factor
driving white views of black leadership was Bradley™s policy record. And
since the majority of whites thought the city had fared well under Bradley,
Bradley™s record was critical to his ongoing success.15
Whites who were interviewed in greater depth in the years after Bradley
entered of¬ce echoed this transformation and in so doing highlighted
the important role that information played. One white resident of Los
Angeles put it this way: “A lot of people were very suspicious and fearful
before Bradley got in. But they never say anything now. I™m sure they
have changed their opinions. . . . Most important, he is a good person.
Whether he is black or white is immaterial” (boatshop worker interviewed
by the US News & World Report 1975). Mark Murphy, the editor of the

15 Additional analysis suggests that Bradley™s record on race was also critical. When a
question asking about whether or not black-white relations had improved over the course
of Bradley™s tenure was added to the incumbent regression in Table 5.3, it was signi¬cant.
All else being equal, whites who felt that black-white relations had improved over the
course of Bradley™s tenure were 21% more likely to approve of Bradley™s mayoralty than
those who thought that race relations had deteriorated.
Black Mayoral Leadership in Los Angeles 119

Los Angeles Times, expressed a similar opinion: “Tom Bradley is beyond
black in the eyes of most of the people of L.A. Most of us think of him
simply as our mayor” (US News & World Report 1975).
Bradley was also well aware of this change in white perceptions. Indeed,
he felt that it was one of the greatest accomplishments of his career, both
for the city and for the country (Ingwerson 1981): “Race, in my judg-
ment, was not a relevant issue. Never should have been. But it took the
experience of the people to be convinced that it should not and would not
become a factor in how you serve the interests of this city. And having
seen that demonstration, I don™t think anybody can make a case now or
in the future that the color of a candidate™s skin is a factor and should be
of any signi¬cant concern” (Ingwerson 1981).
Pollsters, too, knew that experience with Bradley and black leadership
had transformed white views. In Bradley™s two challenger elections, the
common view among political pundits was that racial concerns about
black leadership were an important liability that prevented many, if not
most, white voters from supporting him. As one prominent pollster put it,
“In essence, Tom Bradley contributed to the [white] backlash sentiment
by being black in a de facto segregated society” (Maullin 1971: 51). But
according to the same pollsters, just a few years later, something had
changed for Bradley the incumbent. Bradley was widely recognized as an
important asset to his liberal coalition. In this revised analysis, Bradley, the
man, was the key to his electoral success. At the end of Bradley™s tenure,
Sonenshein wrote: “In the broadest sense, a major resource of the coalition
has been Bradley™s popularity. This public appeal is the product of his style
and of the meaning attached to having a successful Black mayor” (1993:
185). Similarly, in an article in 1989 that marveled at Bradley™s ability
to avoid criticism and gain reelection, Schwada concluded, “Because he™s
politically bland and non-threatening, he can move in all circles” (1989:
102). Bradley™s tenure and the information it provided had transformed
black leadership from being a major threat to the white community into
something that sparked little to no negative reaction.
The lessons from this study of Los Angeles™ mayoral politics, then, are
fairly clear. Racial learning did occur in Los Angeles. White support for
Bradley was never absolute. Many white residents never voted for Tom
Bradley, and a large number never trusted him or black leadership. But
many white residents did change their minds about Tom Bradley and black
leadership. Just as the information model predicts, experience with black
mayoral leadership appears to have led to a fundamental transformation
of white views and a greater willingness to support the black incumbent.
Changing White Attitudes

After living for years under Tom Bradley, white Angelenos™ uncertainty
about black leadership faded and was replaced by a much more reasoned
and positive assessment of black leadership and the gradual diminution
of a black-white divide in the mayoral politics in the city. By enacting a
fairly racially neutral agenda that helped ensure ongoing white prosper-
ity, Bradley had proven to many whites that black leadership was worth
At the same time, it is important to note that by focusing exclusively on
mayoral politics under Bradley this chapter has obscured three potentially
important trends. First, by focusing on Tom Bradley™s words and actions,
this chapter has ignored a range of factors outside of black leadership that
could have contributed to changes in the white vote in Los Angeles. Over
this time period, there were, for example, fairly dramatic changes in the
racial demographics of the city and the state. If rapid Latino and Asian
American immigration represented more of a threat to whites than did
the presence of a relatively small and stable black population, then grow-
ing support for Bradley might signal less about acceptance of blacks and
more about growing fears of Latino and Asian American power.16 Other
trends, such as the migration of white voters into the city and the state,
continued economic growth across the state, and variations in the ideo-
logical leaning of the state™s voters, could also have in¬‚uenced the white
vote. Although racial learning seems to have played an important role, it
is not at all clear that it was the only factor affecting the white vote in Los
Angeles over this period.
A second missing feature of this chapter is a discussion of race rela-
tions outside the arena of mayoral politics. This is an important omis-
sion because an exclusive focus on mayoral politics obscures consider-
able racial discord in the city. Black mayoral leadership may have had an
important effect on white political behavior, but it is not clear how far
that racial learning process extended. Even a cursory examination of race
relations in the city reveals that black representation is no panacea for
the array of racial con¬‚icts facing the city, and the nation. Bradley™s initial
victory appears to have lessened racial tensions across the city, and for
a decade few instances of severe racial con¬‚ict erupted. But subsequent
developments reveal ongoing racial divides in the city. The 1992 riots
are the most obvious example of the continued importance of race in

16 The fact that whites in California used direct democracy to target Latinos more than
blacks or any other group over this period lends some credence to this theory (Hajnal,
Gerber, and Louch 2002).
Black Mayoral Leadership in Los Angeles 121

Los Angeles, and there have been many other episodes of racial con¬‚ict “
some of which have repeated old patterns, and others that have demon-
strated new tensions (Fears 1998; Gold 2001; Newton 1998).17 It is clear
that whites, blacks, Latinos, and Asian Americans continue regularly to
disagree over how the city™s resources should be allocated. In the end,
experience with black leadership is unlikely to be able to alter the under-
lying interests of any of these groups. If blacks, whites, Latinos, and
Asian Americans truly want different policies, minority representation
will do little to end racial con¬‚ict. If, however, members of these groups
want the same things from their government, then minority representa-
tion may help them to recognize their similar interests. Thus, the lesson
from Los Angeles™ recent history is an optimistic one only in a very limited
Finally, a focus on black leadership overlooks important develop-
ments in Asian American and Latino leadership in the city. The unsuc-
cessful mayoral candidacies of Michael Woo in 1993 and Antonio
Villaraigosa in 2001 and the successful mayoral bid of Villaraigosa in
2005 highlight a range of new and extremely variable racial divisions.
Woo™s candidacy resulted in a fairly strong inter-minority coalition and
seemingly widespread white opposition to Asian American leadership.18
Villaraigosa™s 2001 bid pointed to a sharp black-Latino divide and only
somewhat less widespread white opposition to minority leadership.19 And
Villaraigosa™s 2005 electoral victory resulted in a much more moderate
black-Latino divide and sharply increased white support for the Latino
candidate. In Chapter 7, I will look more closely at these new forms of
minority leadership. However, it may be too early to offer much in the
way of de¬nitive generalizations about how Asian American and Latino
leadership will ultimately affect racial dynamics in Los Angeles and other
cities. The mixed results we see in these three elections and the, as of yet,
small number of cases of Latino and Asian American leadership make it
dif¬cult to reach any conclusions.

17 The increasingly central battle between blacks and Hispanics over each group™s political
representation in the city is perhaps the most important new facet of the city™s racial
18 Woo captured 69% of the Asian American vote, 86% of the black vote, and 57% of the
Latino vote but only 33% of the white vote (Kaufman 1998). At least some of this white
opposition could, however, have been generated by Woo™s strongly liberal stances on a
range of issues.
19 In the 2001 primary, according to a Los Angeles Times poll, only 26% of white voters
supported either of the two Latino candidates and in the general election, 59% of white
voters in the city opposed Villaraigosa.
Changing White Attitudes

Another question this study of Los Angeles cannot answer is why some
cities do not conform to the same pattern of growing white acceptance of
black leadership. As we saw in Chapter 4, in a small minority of cities,
whites respond to black leadership with ongoing white opposition and
persistent racial tension. In Chapter 6, I follow the course of black mayoral
leadership in Chicago to try to understand why whites in racially balanced
cities do not change their minds about black leadership.

Black Mayoral Leadership in Chicago

Although black representation in most cases leads to decreased racial
tension and greater acceptance of black incumbents, there are a select
number of cities where racial tension remains high, voting continues to
be highly racially polarized, and few new white voters begin to support
black leaders despite years under black leadership. The goal of this chapter
is to look much more closely at the course of black leadership and white
learning in one such city to try to understand and explain the lack of
Chicago represents perhaps the most famous case of ongoing white
resistance.1 Voting ¬gures in the city tell a story of unrelenting white oppo-
sition to black leadership. Harold Washington actually lost white support
when he ran for reelection and none of the eight black candidates who
have sought the mayoralty after Harold Washington™s tenure has managed
to garner more than 10 percent of the white vote in a primary or general
election. Whites in the city seem no more willing to support black mayoral
leadership today than they were in the days before Washington assumed
of¬ce. The obvious question is why has experience with black leader-
ship produced little, if any, change? Why didn™t white voters in Chicago
begin to accept and support black mayoral leadership as they did in other

1 Another important reason to focus on Chicago is the extensive empirical record on may-
oral and racial politics in Chicago. A range of primary accounts and a number of in-depth
polls detail white attitudes and views before, during, and after Harold Washington™s
tenure as mayor allowing for a detailed assessment of the information model and its

Changing White Attitudes

In Chapter 1, I suggested that information and control were the critical
variables explaining variation across cities “ in particular, in accounting
for the lack of change in a small number of racially balanced cities, such
as Chicago. Racially balanced cities are different, I argued, for two rea-
sons. First, in racially balanced cities, black incumbency provides little
information to white residents because it is unclear whether the black or
the white community is in control of the local political arena, and as a
result it is unclear who should be blamed or credited for conditions in the
city. Rather than experiencing the effects of a black mayoralty, white resi-
dents are apt to witness a battle for control between black and white elites
that ends in stalemate. Second, the precarious balance of power between
the black and white communities in racially balanced cities means that
any additional black victory takes on added signi¬cance, as any black
victory could lead to a black takeover. The limited information provided
by experience with black leadership and the impending possibility of real
black control should, according to the information model, prevent white
support for black leadership from growing and in some cases even inspire
whites to increase their efforts to prevent a black takeover.
To see if this information story can account for the lack of change in
white political behavior in Chicago, I will address a series of questions
about three different stages in the racial learning process. First, did white
fears about the possibility of black leadership in Chicago mirror white
concerns in other cities? Second, how does the information provided by
black leadership in the city of Chicago differ from the information pro-
vided by black leadership in most other cities? In particular, did Harold
Washington™s tenure as mayor provide whites with less credible evidence
about the consequences of black leadership? Third, is a lack of informa-
tion re¬‚ected in ongoing white fears and continued white resistance to
black leadership?
This review of these different stages in Chicago suggests that a lack of
credible information can plausibly account for ongoing white resistance
in the city. Harold Washington™s tenure provided much less information to
white residents than other cases of black leadership because Washington
was prevented by a white-led coalition in the city council from enacting
his agenda. Since a black agenda was never realized, white residents did
not know what would happen if a black mayor actually got control of the
city. The situation was exacerbated by the fact that blacks were moving
closer to achieving control of the city™s political apparatus. Toward the
end of Washington™s tenure, his coalition won two seats on the city council
to create an even 25:25 split. Another loss at that point could easily have
Black Mayoral Leadership in Chicago 125

pushed the precarious balance of power between the black and white
communities in favor of the black community. In this context, it made
sense to many whites in Chicago to continue to oppose black mayoral
As plausible as the information account may be, lack of information is
certainly not the only factor that could have contributed to enduring white
opposition in Chicago. Later in this chapter, I will examine a number of
other factors related to the electoral context that could have in¬‚uenced the
white vote. Most of these factors cannot easily account for ongoing white
resistance but it is impossible with just one case to rule out all factors.
Thus, the conclusions of this chapter at best will be tentative.

concerns about black leadership
To see if information “ or more speci¬cally a lack of information “
could have contributed to ongoing white aversion to black leadership
in Chicago, it is important to establish that whites in Chicago felt many
of the same fears that other whites in other cities felt when faced with
the prospect of black mayoral leadership. A review of the sentiments
expressed by white residents, the tactics of the candidates, and patterns
in the white vote all strongly suggest that whites in Chicago were, in fact,
just like whites in other cities. They were uncertain and afraid. Just as
in other cities with serious black challengers, the uncertainty that sur-
rounded black leadership was acknowledged, explicitly or implicitly, by
almost everyone who participated in the election.
White residents were relatively open about their fears and concerns. For
many white residents, the prospect of black leadership implied nothing
short of disaster. “[It™s] more than an election,” one voter opined. “We™re
¬ghting for everything we have this time” (Kleppner 1985). Another
said: “The story is going around here ˜Go in your room and lock your
door for two days if Washington is elected™”(Peterson 1983b). “If Harold
Washington is elected mayor,” a third city resident predicted, “it will be
the worst disaster in Chicago since the Chicago ¬re” (quoted in Rivlin
1992: 191).
Often, the fear was generalized. If a black man is elected, bad things
will happen. But frequently, it focused on one terrifying possibility: neigh-
borhood integration. As one white resident put it, “I don™t know how to
say this, but most people are afraid he is going to exert all of his powers
for the black community and the white community is going to get nothing.
My fear is that he™s going to try to push racial integration, which is ¬ne
Changing White Attitudes

as long as I don™t lose money on my house . . . because I can™t take the loss
(Coleman 1983a). A Democratic Committee member concurred, explain-
ing, “[My constituents] are giving me a message of racial pride . . . They™re
afraid of scattered-site housing. They™re concerned about the stability of
our neighborhoods” (in Rivlin 1992: 185“6). “People have lived here all
their lives,” a resident of one white neighborhood explained. “It™s a nice
area and they want to hold onto what they have” (Morganthau 1983).
Jesse Jackson put it more succinctly and perhaps more accurately than
anyone else: “Black people were energized. White people were trauma-
tized” (quoted in Kleppner 1985: 187).2
As in other cities facing the possibility of black leadership, the white
candidate attempted to play on these racial concerns. “Epton, before it™s
too late!” became the rallying cry of Washington™s white opponent in
the general election. Mirroring Sam Yorty™s racialized campaigns in Los
Angeles, Bernard Epton asked white voters to consider what would hap-
pen if a black man were elected to run the city: “Will he obey the law?
Will he do what he promises?” (Kleppner 1985: 205). And, as in other
cities, the black candidate had to respond. Washington, while primarily


. 4
( 7)