. 1
( 7)


This page intentionally left blank
Changing White Attitudes toward Black Political Leadership

Despite the hopes of the civil rights movement, researchers have found
that the election of African Americans to of¬ce has not greatly improved
the well-being of the black community. By shifting the focus to the
white community, this book ¬nds that black representation can have
a profound impact. Utilizing national public opinion surveys, data on
voting patterns in large American cities, and more in-depth studies of
Los Angeles and Chicago, Zoltan L. Hajnal shows that under most
black mayors there is real, positive change in the white vote and in the
racial attitudes of white residents. This change occurs because black
incumbency provides concrete information that disproves the fears and
expectations of many white residents. These ¬ndings not only highlight
the importance of black representation; they also demonstrate the crit-
ical role that information can play in racial politics and point to the
ability of at least some whites to change their minds about blacks and
black leadership.

Zoltan L. Hajnal is an assistant professor of political science at the
University of California, San Diego. He received his Ph.D. in political
science from the University of Chicago. He has published articles in
numerous journals, including the American Political Science Review, the
Journal of Politics, Urban Affairs Review, and Social Science Quarterly.
He received the American Political Science Association™s award for Best
Paper on Urban Politics. His research has been funded by the Russell
Sage Foundation and the Andrew W. Mellon Foundation.
Changing White Attitudes toward Black
Political Leadership

University of California, San Diego
cambridge university press
Cambridge, New York, Melbourne, Madrid, Cape Town, Singapore, São Paulo

Cambridge University Press
The Edinburgh Building, Cambridge cb2 2ru, UK
Published in the United States of America by Cambridge University Press, New York
Information on this title: www.cambridge.org/9780521857475

© Zoltan L. Hajnal 2007

This publication is in copyright. Subject to statutory exception and to the provision of
relevant collective licensing agreements, no reproduction of any part may take place
without the written permission of Cambridge University Press.

First published in print format 2006

978-0-511-26008-7 eBook (EBL)
isbn-10 0-511-26008-3 eBook (EBL)

978-0-521-85747-5 hardback
0-521-85747-3 hardback

0-521-67415-8 paperback

Cambridge University Press has no responsibility for the persistence or accuracy of urls
for external or third-party internet websites referred to in this publication, and does not
guarantee that any content on such websites is, or will remain, accurate or appropriate.

Acknowledgments page vii

Introduction 1
1 Black Leadership: The Possibilities 14
2 The Transformation of the White Vote 38
3 The Transformation of White Attitudes 73
4 Learning Across Different Cities 91
5 Black Mayoral Leadership in Los Angeles 103
6 Black Mayoral Leadership in Chicago 123
7 Other Cases Where Information Could Matter 141
Conclusion: A Tale of Caution and Hope 159

Statistical Appendixes 169
A: Appendix to Chapter 2 171
B: Appendix to Chapter 3 177
C: Appendix to Chapter 5 182
References 189
Index 213


As with most other academic endeavors, this book has been a long time
coming. It started as an off-hand comment early in graduate school and
has grown through graduation, fellowships, and my time as a faculty
member at the University of California, San Diego. Over all those years
I have incurred countless intellectual and emotional debts. If so many
people had not taken the time and energy to help me along, this project
would de¬nitely have suffered in quality and it may never have been
completed. I can never begin to repay all of the debts, but I would at least
like to acknowledge those who helped make this book much better than
it otherwise would have been, and perhaps even more importantly those
who helped keep me sane through an often arduous process.
I want to ¬rst say thanks to my friends who labored with me in graduate
school. John Baughman, Greg Bovitz, Pam Cook, Nancy Crowe, Robert
Eisinger, Kevin Esterling, Anna Greenberg, Karen Hoffman, Thomas Kim,
Roger Larocca, and Chris Parker all offered helpful advice and, on occa-
sion, unwelcome but very necessary criticism. The rest of the University
of Chicago American Politics Workshop, where I presented much of this
research, also deserve thanks. Four of my partners in crime, Paul Frymer,
Jamie Druckman, Andrew Grant-Thomas, and Taeku Lee, deserve special
recognition. Each provided seemingly endless support; their comments
and kindness have had a profound in¬‚uence not only on the end product
but also on me as a person.
I owe large and lasting debts to my dissertation committee, Michael
Dawson, Don Green, William J. Wilson, and Lynn Sanders. Michael read
every word of every draft and although I didn™t always like to hear what he
had to say, this would have been a much less worthwhile project without

his assistance. Don Green, who was my ¬rst true mentor and who I still
hold responsible for getting me mixed up in the business of political sci-
ence, has been there all along. His enthusiasm and encouragement were
invaluable and his scholarly advice equally indispensable. William J. Wil-
son™s contributions were also numerous. Above and beyond his counsel,
I bene¬ted from his help in getting an Urban Poverty Fellowship, which
not only provided me with a stimulating environment in which to work
but also helped to pay for dinner.
After leaving Chicago, I had the good fortune to spend time at four
different institutions, Brandeis University, the Public Policy Institute of
Chicago, the University of California, San Diego, and Princeton Univer-
sity, each of which provided a stimulating intellectual home. Along the
way, I met a number of close colleagues who generously gave of their time
and advice. Sidney Milkis at Brandeis was instrumental in giving me the
con¬dence to transition from student to scholar. Debbie Reed, Paul Lewis,
and Mark Baldassare at PPIC gave me the time and resources to complete
my American Political Science Review article “ an important building step
in the intellectual development of this project. My colleagues at UCSD,
Amy Bridges, Steve Erie, Karen Ferree, Clark Gibson, Peter Gourevitch,
Gary Jacobson, David Lake, Mat McCubbins, Phil Roeder, Sam Kernell,
Thad Kousser, and Sam Popkin, helped me to formulate my ideas and
work through the all-too-frequent problems. Their insights and advice
have shaped the book in a range of important ways. Among my UCSD
colleagues, Amy Bridges stands alone. Over the years here, she has been
an incredible mentor and friend.
I am also very grateful to the numerous academics who gave of their
time not because they had to but simply because they cared. Charles Bul-
lock, Ira Katznelson, Arthur Lupia, Tony Marx, and Raphael Sonenshein
all read and commented on different sections of the book. Paul Snider-
man™s sage advice was critical. He has left a profound mark on the ¬nal
product. I also owe a special debt to Liz Gerber, who over the years has
not just read my work but has encouraged my efforts and offered valu-
able assistance on both a professional and a personal level. Other scholars
associated with the Center for Advanced Behavioral Studies at Stanford,
the Institute for Governmental Studies at UC-Berkeley, and the American
Politics Workshop at Harvard also gave their time and provided important
I am proud to be associated with Cambridge University Press. Dennis
Chong and Jim Kuklinski paved the way for me at the Press and then
handed off the project to Lew Bateman, who has done an excellent job
Acknowledgments ix

of guiding this manuscript from rough draft to more polished product. I
thank all three of them for believing in this book. Anonymous reviewers
at the Press similarly deserve thanks for forcing me to present clearer and
more compelling evidence.
I would be remiss not to mention the money that made it all happen.
I was lucky enough to get support for this project from the University
of Chicago, the Mellon Foundation, the National Science Foundation (in
the form of an Urban Poverty Fellowship), the Center for the Study of
Democratic Politics at Princeton University, and the Hellman Fellowship
at UCSD. For all this support, care, and advice, I am amazed and grateful.
Anyone who knows me knows that my family is my rock. My parents,
Vivian and Zoltan, and my sister, Catherine, were always there for me on
the far too many days when I was down and defeated. They picked me
up, pointed me forward, and gave me all the love I needed to continue.
From my family, I have learned not only to endure but how to live. I offer
them my deepest, warmest thanks. They will always be front and center
in my heart.
In the middle of this project, a new little girl came into my life. And
already by age four, Lina has been able to offer invaluable assistance. As
a true native Californian, when times were tough, she simply told me,
“Don™t worry daddy. Relax. Breathe in, breathe out.” It seemed to work.
Lina is often a distraction from work but she is a distraction that I will
cherish forever.
Last and clearly not least, I offer my thanks, my friendship, and my
love to my wife, Barbara Walter. Over the past years she has played every
imaginable role in my life and in my career. Without her I would not have
begun this project. Without her I certainly would not have ¬nished. Her
words of advice, her love, and “ yes “ her thoughtful criticism helped me
more than I can say. For that and much, much more I dedicate this book
to her.

One of the great hopes of the civil rights movement was that African
Americans, by gaining the right to vote, would be able to elect represen-
tatives who could ultimately reduce or even eradicate racial inequality. To
many in the community, black elected of¬cials were “saviors who were
going to uplift the people, eradicate police brutality, house the homeless,
[and] ¬nd new jobs for everyone who was struggling.”1 In the 1960s,
1970s, and 1980s, as blacks began to win of¬ce and displace thousands
of white incumbents, many in the African American community were
understandably jubilant. As one voter who witnessed the transition put
it, “It was almost like the feeling you have when you see your ¬rst-born “
a sense of accomplishment, of utter elation” (quoted in Donze 1998).
Decades later, it is clear that black representation has made a differ-
ence. Many black leaders have tried valiantly to improve the lives of their
black constituents, and black representation at different levels of of¬ce
has been associated with concrete, positive change for the black commu-
nity. It has led directly to increases in the numbers of African Americans
in many city governments (Browning, Marshall, and Tabb 1984; Eisinger
1982; Levine 1974), to greater black political participation (Tate 2003;
Gay 2001; Bobo and Gilliam 1990), to modest shifts in spending policies
(Brown 1996; Karnig and Welch 1980), and to the implementation of
reforms to police practices (Headley 1985; Lewis 1987).2 But none of these

1 Bill Campbell, former mayor of Atlanta, made this comment regarding the expectations
surrounding the ¬rst black mayors (quoted in Fulwood 1995).
2 Not everyone agrees that black leaders have made a difference. Some scholars have argued
that black incumbents have done no more for the black community than white incumbents

Changing White Attitudes

changes has been dramatic. According to most studies, black political
representation has not lived up to expectations (Smith 1996; Singh 1998;
Reed 1988; Perry 1990; G. Peterson 1994; Browning, Marshall, and Tabb
1997; Marable 1992). Despite large gains in the number of black elected
of¬cials across the country, there has been only moderate change in basic
indicators of African American well-being and, even more importantly,
almost no change in various measures of racial inequality.3 Though black
of¬cials have controlled the mayoralty in seven of the ten largest cities
in the country and have achieved nearly proportionate representation in
the House of Representatives, ¬gures comparing black to white poverty,
unemployment, and educational attainment remain largely unchanged. In
1967, when the ¬rst big-city black mayors were elected in Cleveland and
Gary, blacks were three times more likely than whites to be poor, twice
as likely to be unemployed, and one-third as likely to have completed
college. Today, with more than 9,000 black elected of¬cials across the
country, those ¬gures are nearly identical (Blank 2001; Dawson 1994).
Richard Arrington, mayor of Birmingham for twenty years, summed up
the situation when he was asked what blacks had to show economically
for his tenure in of¬ce: “Quite frankly,” he said, “we don™t have very
much” (quoted in Edds 1987).
But only part of the story of black political representation has been
told. Studies have overlooked important gains associated with black

in similar cities and districts (Mladenka 1989, 1991). Swain (1995), for example, has
argued that white Democrats have done as much for African American interests as black
Congress members from the same types of districts (but see Tate 2003; Whitby 1998;
Herring 1990).
3 The black community experienced undeniable gains in the early and mid-twentieth century
(Thernstrom and Thernstrom 1997). The black middle class, for example, grew from just
12% of the black population in 1949 to 41% today (Farley 1996). But since the late 1960s,
the story has grown more complicated. Blacks have made progress in absolute terms. High
school graduation rates, for example, have improved considerably, and earnings have
increased slightly (Farley 1996). But relative to whites, most indicators of black well-
being reveal little change in the past several decades (Blank 2001; Klinkner and Smith
1999; Thernstrom and Thernstrom 1997). Despite the passage of civil rights legislation,
increased social interaction between whites and blacks, and some claims that race has
been diminishing in signi¬cance in recent decades, studies reveal only minimal decreases in
residential segregation (Massey and Denton 1993; Massey 2001), fairly widespread racial
discrimination in hiring and the housing industry (Kirschenman and Neckerman 1991;
Bertrand and Mullainathan 2004; Massey and Denton 1993), persistent racial stereotyping
(Bobo and Johnson 2000; Lee 2000), and strong racial undertones to many of the political
choices whites make (Mendelberg 2001; Carmines and Stimson 1989; Gilens 2001). In
short, there are few signs of major gains since 1970 and plenty of evidence that race retains
much of its signi¬cance in American life (Dawson 1994).
Introduction 3

of¬ceholding because they have focused almost exclusively on the black
community while essentially ignoring the white community. In this book,
I explore how experience with black leadership affects the attitudes,
actions, and political choices of white Americans. Examining white reac-
tions to black leadership “ looking speci¬cally at changes in the racial
attitudes, voting behavior, and policy preferences of white Americans “
demonstrates that black representation has meaningful and positive
effects that are rarely considered in evaluations of the performance of
black leaders. Although the election of African Americans to public of¬ce
has not yet improved the condition of blacks to the degree many people
had hoped, it has had a signi¬cant impact on white attitudes and voting
behavior, and these shifts, though small, could ultimately be the catalyst
for the acceptance of more signi¬cant progress toward racial equality in
American society.

the information effects of black leadership
Experience with black incumbents has real consequences for many mem-
bers of the white community because it imparts critical information about
black preferences that reduces whites™ uncertainty and fear about blacks
and black leadership; this information essentially changes the way that
many white Americans think about the black community and therefore
subtly alters the nature of racial politics and race relations in this country.
Prior to the election of a black candidate, most white voters have little or
no experience with black leadership. For this reason, many rely on racial
stereotypes and past patterns in race relations to assess the likely con-
sequences of a black candidate™s victory. The result is that many whites
fear that a black leader will favor the black community over the white
community. They expect a black leader to redistribute income, encourage
integration, and generally channel resources toward the black community.
In short, they imagine that black control will have negative consequences
for themselves and their neighbors. Once a black candidate is elected,
however, whites gain access to better information about the policy pref-
erences of black leaders and the effects of black leadership. They become
able to judge black candidates on their records. And because the white
community rarely suffers under black incumbents, those records are, in
almost every case, better than white stereotypes and fears suggested they
would be. When blacks have the power (or are perceived as having the
power) to in¬‚ict harm on the white community and they choose not to
do so, many whites are forced to reevaluate their assumptions.
Changing White Attitudes

The idea that white behavior in biracial electoral contests is governed
by uncertainty and information is a novel one. Existing explanations of the
voting behavior and attitudes of whites tend to focus on two very different
mechanisms: (1) prejudice and (2) white backlash against perceived racial
threat. Taking these in order, it has been argued that black representation “
no matter how positive its effect on the white community “ should have
little or no effect on white attitudes and political behavior because white
Americans are basically prejudiced and unwilling or unable to change
their views of blacks (Hurwitz and Pef¬‚ey 1998; Kuklinski et al. 1997;
Allport 1954; Adorno et al. 1950). If prejudice is indeed behind white
opposition to black empowerment, then the words and actions of black
incumbents cannot affect whites™ views, because these views are too stable
and too deeply ingrained to be easily altered (Fazio et al. 1995; Devine
1989; Fiske 1998; Rothbart and John 1993). And even when whites expe-
rience black leadership and gather information from the experience that
runs counter to stereotypes and expectations, they will simply ignore or
discount evidence that challenges their prejudices (Macrae, Hewstone,
and Grif¬th 1993; Weber and Crocker 1983; D. Hamilton, 1981).4 The
second model suggests that black leadership spurs white backlash. At
least one text (Sidanius, Devereux, and Pratto 1991) argues that whites
have a strong incentive to protect America™s racial hierarchy and their
hegemonic position within it. Indeed, past patterns in race relations indi-
cate that when white social status is threatened by black gains, mem-
bers of the white community tend to react by mobilizing to reverse those
gains (Olzak 1990; Stenner 1995). If past patterns prevail, the election
of blacks to of¬ce might represent just another step in an ongoing racial

understanding the variation in reactions
Certainly, neither of these two theories applies to all white Americans
across all contexts. Why do whites in some cities learn to accept a black

4 The fact that stereotypes of blacks are still widespread is taken by many as evidence
that “blatantly prejudiced attitudes continue to pervade the white population” (Kuklinski
et al. 1997). The speci¬c terms that whites use to describe blacks may have changed but
there is ample evidence that large segments of the white community continue to see blacks
as less intelligent, less hardworking, more dif¬cult to get along with, and more violent
than whites (Bobo and Johnson 2000; Lee 2000; Devine and Elliot 1995; Schuman et al.
Introduction 5

mayor, while in other cities whites™ opposition remains constant or even
grows? And why within a particular city do some white residents react
more positively to black leadership than do others? In addition to assessing
the general nature of white reactions to black leadership and testing my
information model of those reactions against the existing prejudice and
white backlash models, it is an important secondary goal of this study to
explain variation in whites™ reactions to black leadership.
Variations in white reactions follow predictable patterns. First, white
reactions are affected by the actions of speci¬c black leaders and the infor-
mation that those actions provide. A black mayor who presides over a city
where housing prices plummet and crime soars is likely to provide white
residents with different information, for example, than a black mayor who
aids in a city™s renaissance. But the actions of particular leaders are not
the central factor governing white reactions, because black representation
almost always proves to be less detrimental to white interests than many
whites fear. What accounts for most of the variation in white responses is
not variation between individual black leaders, but rather white voters™
judgment of the credibility of the information that they receive about black
of¬ceholders: the more power that whites believe black leaders have, the
more they will credit and be in¬‚uenced by the information they receive
from those leaders™ words and deeds. Practically speaking, this means that
whites™ reactions to black representatives are heavily dependent on racial
demographics, which in¬‚uence a representative™s ef¬cacy in of¬ce. In addi-
tion to variation among leaders and across locations, it is important to
consider differences between individuals, focusing particularly on parti-
sanship and exploring the question of whether Democrats or Republicans
are more likely to learn from black leaders.

why black representation and white learning matter
Understanding the relationship between black leaders and white voters is
important for a number of reasons, both substantive and theoretical. It is
clear from the trends in the number of black elected of¬cials that African
American representation is an important and growing phenomenon. In
1960 only 280 blacks held of¬ce across the entire United States (Jaynes
and Williams 1989). Today there are over 9,000 black elected of¬cials
in America (JCPS 2003). Blacks have won the mayoralty in most of the
nation™s big cities, there are roughly 600 African Americans in state leg-
islatures nationwide, and blacks now hold about 10 percent of the seats
Changing White Attitudes

in the U.S. Congress. African Americans are still underrepresented at most
levels of government, but undeniably they play a role among America™s
political elite.
Moreover, white voters are becoming increasingly critical to black elec-
toral victories. Each year more blacks win of¬ce in racially mixed and
predominantly white areas (Bositis 2002). Already six of the ten largest
plurality white cities have had black mayors. Douglas Wilder™s term as
governor of Virginia marked the ¬rst time a black politician had been
elected governor of an American state. Notable black congressional rep-
resentatives such as Julia Carson, Robert Scott, and Barbara Lee can also
be added to this expanding list of successful cross-over candidates.
It is also clear that if black representation is to continue to expand,
black candidates will have to win over more white voters. Black politicians
already represent most of the majority black districts and cities around
the country (Handley and Grofman 1994; Handley, Grofman, and Arden
1997). In addition, court decisions in the 1990s have made it more dif¬cult
to alter electoral lines to create additional majority-minority districts. If
more blacks are to be elected, they will have to win in racially mixed
Black representation, furthermore, may be setting the trend for an
even bigger phenomenon: Latino and Asian American representation.
The Latino population is expected to double in the next ten years. By
mid-century, Latinos may represent as much as one-third of the U.S.
population, while Asian Americans, currently the fastest-growing pop-
ulation in the country, could account for almost 10 percent (Bureau of
the Census 2002b). Latino and Asian American representation still lags
far behind African American representation, but these demographic pro-
jections suggest that the situation may change relatively quickly. Already,
recent gains in Latino and Asian American of¬ceholding have far out-
stripped black advances (NALEO 2002; APALC 2003). What all of this
suggests is that minority representation is likely to become an increasingly
central aspect of American politics. Whether white and non-white Amer-
icans follow a path toward mutual understanding and interracial coop-
eration or move instead toward distrust and escalating con¬‚ict may well
depend upon today™s minority leaders and their interactions with white
In addition to speaking to these substantive issues, this study provides
insight into a number of important theoretical questions about the nature
of race and politics in America. One of the most central debates in Amer-
ican politics today concerns how much race shapes political choices. On
Introduction 7

one side of the debate are scholars who insist that race and racial pre-
judice remain the primary factor in American politics in general and in
white voting preferences in particular (Reeves 1997; Bell 1992; McCrary
1990; Huckfeldt and Kohfeld 1989). According to these scholars, “racism
is an integral, permanent, indestructible component of this society” (Bell
1992: 217). Epitomizing this camp, Robert Starks maintains that “race
is such an overriding factor in American life that to support its elimina-
tion or diffusion as a factor in elections through deracialization is folly”
(1991: 217). On the other side of the debate stand those who believe that
race has lost much of its signi¬cance in the electoral arena and that white
voters are now willing to support black candidates in greater numbers
(Swain 1995; Thernstrom and Thernstrom 1997). As Abigail Thernstrom
notes, “Whites not only say they will vote for black candidates; they do
so” (Thernstrom 1995). Some scholars even suggest that race is no more
an issue in biracial elections than it is in other electoral contests (Highton
2004; Citrin Green, and Sears 1990; Thernstrom 1987). A black candi-
date is likely to lose, they argue, for many of the same reasons that a white
candidate is likely to lose.
One of the central goals of this book is to show that this debate
addresses the wrong issue. The key question is not if race is central in
the minds of white voters, it is when race is central in the minds of white
voters. By showing that the transition from white to black leadership
frequently leads to notable shifts in white attitudes and behavior, I will
demonstrate that race plays a much more dynamic role in American pol-
itics than we have understood. Though race and racial prejudice remain
prevalent in American society, change is possible under the right circum-
stances. To really understand how race “works” in the American context,
we have to ¬nd out when racist voting is more likely, when color-blind
politics tend to emerge, and ultimately why these differences occur.
In this study, I also make important observations regarding the role
that information plays in the minds and voting decisions of the Ameri-
can population. For decades, scholars have argued that Americans simply
do not have enough information about politics to make reasoned, ratio-
nal decisions (Campbell et al. 1960; Converse 1964; Delli Carpini and
Keeter 1996). It is true, for example, that less than half of all Americans
know both the name and the party af¬liation of their representative in
Congress (Jacobson and Kernell 1981). Many cannot even distinguish
between the policy platforms of the Democratic and Republican parties
(Bennett 1995). What political knowledge Americans do have is usually
not molded into coherent, consistent reasoning about issues and events
Changing White Attitudes

in the electoral arena (Converse 1964).5 From this viewpoint, it would
seem unrealistic to expect experiences under a relatively small number of
black representatives to inspire real change in the views or actions of the
public. But recent scholarship suggests that the average American does
have enough information to make reasonable decisions about the polit-
ical arena (Lupia 1994; Popkin 1991; Lupia and McCubbins 1998; Lau
and Redlawsk 1997). If the job of evaluating leaders only requires indi-
viduals to know basic facts about their own well-being and trends in the
welfare of their communities, then they may have enough information
in their daily personal lives. Moreover, there is clear evidence that voters
regularly incorporate current events in the making of political decisions
(Popkin 1991; Alvarez 1997; Bowler and Donovan 1994; C. Franklin and
Jackson 1983; Allsop and Weisberg 1988).
In keeping with this recent trend, one of my central contentions is
that politics “ even local politics “ can be extremely informative and
consequential. Under the right circumstances “ for the purposes of
this study, when race is involved “ Americans will pay attention to the
political arena and will assess local politicians by evaluating conditions
in their own communities. Moreover, this evaluation can have real
consequences. By showing that whites tend to oppose black challengers
when they are uncertain about how black leadership will affect them, but
that they become measurably more willing to support black incumbents
when they have experienced black leadership and know more about
its effects on their well-being, I hope to con¬rm the critical role that
information plays in the arena of racial politics.
Finally, there are obvious implications for how we view descriptive
representation and the degree to which we should try to expand minority
representation. If the “politics as usual” that frequently occurs when black
representatives are elected has a positive impact on white Americans and
leads to a change in the white vote and in the racial sentiments expressed
by a sizeable part of the white electorate, then there is at least one reason
to try to expand descriptive representation. And if black leaders can help
black constituents “ even if only to a limited extent “ while at the same
time subtly changing white views and votes, this alone would seem to

5 As a result, many claim that political decisions are predominantly shaped by long-term
forces, such as party identi¬cation, which are acquired early in life and are not easily
changed (Campbell et al. 1960; Beck and Jennings 1991; Green et al. 2002; Green and
Palmquist 1990).
Introduction 9

make it imperative to create a body of elected of¬cials that more closely
resembles the public.

the empirical strategy
The empirical goals of this book are twofold. The ¬rst goal is to offer a
broad account of how white Americans react to having African Americans
as their leaders. Few researchers have even thought to ask about the impact
of black leadership on the white community. Fewer still have tried to
answer that question. And no one has answered it in a systematic way.
Studies that have touched on the relationship between black leaders and
their white constituents have been largely anecdotal in nature “ focusing
on one leader or city “ and often limited in their scope “ focusing on
only one aspect of white behavior. The result is a range of contradictory
conclusions. We simply do not know how white residents respond to black
To assess how black leadership affects the white community, I will focus
on two critical measures of white political behavior. The ¬rst is the vote.
After experiencing black leadership, are white Americans more or less
likely to support black candidates? The second is racial attitudes. After
experiencing black leadership, are white Americans likely to view blacks
and black leadership more positively, more negatively, or about the same
way? If black leadership can bring about real, positive change on both of
these measures, it is clear that the election of African Americans to of¬ce
represents an important step in American race relations.
For each of these two measures, I will assess changes in white political
behavior as systematically as possible. Rather than examining a single
city or a single leader, I will examine an entire universe of cases of white
reactions to black representation. In particular, throughout the book I
will analyze white reactions across the full range of cases of black may-
oral leadership. That analysis will include an examination of every black
incumbent™s reelection bid in the twentieth century, a comparison of a
complete set of black challenger and black incumbent electoral bids, and
a test of white views across a nationally representative sample of cities.
Once all of these tests have been performed, we should have a complete
and fairly compelling picture of the impact of black leadership on the
politics of the white community.
The other important empirical contribution of this book is to test the
three different theoretical accounts of the white community. Are most
Changing White Attitudes

whites really governed by the information model or is social dominance
or racial prejudice a more important determinant of white reactions to
black leadership? Fortunately, since each of these theories offers different
predictions about changes in the white vote and in white attitudes, by
looking systematically at how black leadership affects the vote and racial
attitudes, we should be able to determine which of these three models is
at play.

the case of black mayors
I focus on the implications of black leadership at the mayoral level for four
reasons. First, in order for white residents to react in any way to black lead-
ership, they must be aware that black leadership exists. Whereas the names
of school board members or lower court justices are relatively unknown
to the residents of most communities, people can usually identify their
mayor. In cities with black mayors, in fact, available evidence suggests
that the overwhelming majority of residents can identify the mayor (Cole
Second, to test the willingness of whites to support black leadership,
the of¬ce to which a black person is elected must be viewed as important
and powerful. Whites may support black candidates who seek of¬ces
that whites perceive as powerless and unimportant without fear of the
consequences. The mayoral of¬ce, which is considered by most people
to be a powerful and in¬‚uential post, represents a truer test of white
willingness to support black leadership than would a test of another, less
powerful position.
Third, for white residents to be able to judge black leaders, they must
be able to observe the actions of a black incumbent and connect them
to changes in local conditions or policies. As the executive of a city, the
mayor focuses on local issues and often acts unilaterally. State and national
legislators, by contrast, often concentrate on regional or national issues
and must generally obtain the support of their colleagues before acting.
Even though the of¬cial powers of mayors are often quite limited, evidence
strongly suggests that the public views them as responsible for local con-
ditions. A poll undertaken in Washington, DC, where the mayor™s power
is limited, found that a clear majority of city residents believed the mayor
“can control” or “exact in¬‚uence” on almost every policy issue facing the
city (Coleman and Sussman 1978: A1). Because residents feel that their
mayor has the power to in¬‚uence policy, they are quite willing to judge
black leadership in general on the basis of a black mayor™s performance.
Introduction 11

Finally, on a more practical level, black mayors make an excellent
choice simply because there are so many of them. In 2001, there were
451 black mayors in the United States, 49 of whom served in cities with
a population of over 50,000 (JCPS 2002). Black mayors have held of¬ce
in four of the ¬ve largest American cities, and, in fact, almost 10 per-
cent of all big-city mayors are African American (JCPS 2002). There are
enough cases of black mayors in cities with different racial demograph-
ics to allow for empirical analysis; the availability of data on city-level
voting patterns, local government spending, campaign rhetoric, and eco-
nomic and social conditions makes the mayoralty one of the only feasible
choices for statistical study.
Differences between the mayoralty and other political of¬ces mean
that the results of this study cannot necessarily be applied to other types
of black leadership. In fact, I believe that white learning is likely to be
more pronounced under black mayors than under most other types of
black leaders. As I suggested earlier, when black people are elected to
serve as state or federal legislators, they are less likely than black mayors
to be viewed as being responsible for local conditions or policies; thus,
experience with black legislators should do less than experience with black
mayors to change white views and votes. By the same token, there may
be of¬ces that foster more pronounced learning. A black president, for
example, would surely be seen as much more powerful than a black mayor
and would therefore present an interesting and important test case of the
information model.

an overview
This study proceeds as follows: the ¬rst chapter explains the three different
theories of black representation “ the information model, the prejudice
model, and the white backlash model “ in greater detail and describes
the predictions they make regarding the effect of black leadership on
white political behavior, including the predictions the information model
makes about variation across leaders, individuals, and cities. The chapter
concludes with an overview of existing research and a description of my
Chapter 2 analyzes changes in the white vote under incumbent black
mayors to answer three questions about black candidates: (1) Are white
voters more or less likely to support black challengers than they are to
support the same black candidates when they run as incumbents? (2) Does
black incumbency at the mayoral level lead to changes in the nature of
Changing White Attitudes

the white vote? In particular, does experience with black leadership teach
voters to pay increasing attention to nonracial factors such as endorse-
ments, candidate quality, and the economy? (3) Does information from
one black incumbent play a role when white voters consider other black
candidates? In other words, is any learning that occurs transferable from
one candidate to another?
Chapter 3 looks at changes in white racial attitudes under black may-
ors. What is it that whites do or do not learn from their experiences
with black leadership? Do they simply learn not to fear a particular black
incumbent, or do their attitudes toward the black community also change?
Here, I utilize responses to the American National Election Study (ANES)
surveys, which have over the past half-century polled a representative
sample of Americans about a range of their political views, to compare
white attitudes before and after the election of a black mayor and to
assess changes in white attitudes as experience with black of¬ceholding
I examine variation in white reactions to black leadership from one
city to another in Chapter 4. Because I contend that the credibility of the
information that black incumbency provides is critically dependent on the
amount of control that blacks are able to exercise once in of¬ce, I look
at changes in white voting behavior between black challenger and black
incumbent elections in twenty-¬ve cities in terms of the racial makeup
of these cities to determine whether demographic differences can explain
why whites seem to learn from black leadership more readily in some
places than in others.
In Chapter 5, I look more closely at the course of black mayoral lead-
ership in a single city, Los Angeles. This case study is designed to provide
more direct evidence of the process of racial learning and to demonstrate
as clearly as possible how information from black incumbency slowly
translates into changes in white attitudes and behavior. Throughout the
chapter, I tie the thoughts and actions of white residents in the city to
the actions of Tom Bradley during his twenty-year tenure as mayor. I
then, in Chapter 6, follow white attitudes and actions in Chicago, a city
where black leadership has had little to no positive effect on white political
behavior, in an effort to understand and explain the lack of change. I show
that ongoing white opposition to Harold Washington and subsequent
black candidates in Chicago is at least in part a function of the inabil-
ity of the city™s black leaders to prove themselves. Because Washington™s
tenure led to political stalemate, white residents could not learn about
Introduction 13

the real consequences of his leadership, and widespread white concerns
about blacks taking control of the local political arena remain.
Chapter 7 considers the generalizability of the information model
beyond the mayoral of¬ce, beyond the African American race, and beyond
race itself. I begin by considering the effects of black representation at the
congressional level. After pausing to look at African Americans™ reac-
tions to black mayors, I offer a brief account of the history of Asian
American and Latino elected of¬cials and discuss some of the similarities
and contrasts between race, religion, and gender in terms of the poten-
tial applicability of the information model. In my conclusion I review
the major ¬ndings of the book and discuss a number of the substantive
and theoretical implications of this research. I outline some lessons for
policy makers interested in either expanding minority representation or
structuring electoral districts to minimize racial con¬‚ict.

Black Leadership
The Possibilities

In a nation that has long been divided by race, the election of black leaders
is of great historic importance. But it is in many ways an uncertain step
with unknown consequences. Black leadership raises both meaningful
possibilities and real risks, especially when African Americans are elected
in racially mixed areas. After winning elections, black of¬cials must lead
communities that are racially diverse and often bitterly divided. How
does white America respond to African American leadership? We have
anecdotal evidence from various cases, but we know very little about the
general pattern and ultimate consequences of black leadership: We don™t
know whether minority political leadership tends to exacerbate or reduce
racial tension, whether black incumbents are more or less successful than
their white counterparts in subsequent elections, or under what politi-
cal, economic, and racial conditions white support can be maintained or
increased over time.
In this chapter, I detail three different accounts of black-white relations
that offer predictions about what might happen after the onset of black
leadership. I begin by presenting in full my information model of white
behavior, which focuses on the information that experience with black
incumbents provides white voters. I then contrast this model with two
conventional theories drawn from the existing literature on race relations
and American politics: the prejudice model and the white backlash model.
Finally, I review existing accounts of white responses to black leadership,
note some of the de¬ciencies of these accounts, and suggest a more sys-
tematic approach to assessing the effects of black representation on white
political behavior.

Black Leadership: The Possibilities 15

the information model
The information model suggests that black leadership should signi¬cantly
change the voting behavior of whites and the way white Americans think
about black candidates because the candidates™ terms impart critical infor-
mation that greatly reduces uncertainty and dispels white fears about
blacks and black leadership. The logic is fairly straightforward. When
black challengers run for of¬ce, many white residents are uncertain about
the consequences of black leadership and fear that black leaders will favor
the black community over the white community, thereby reversing the
racial status quo. To prevent this from happening, large segments of the
white community are apt to mobilize to prevent a black electoral vic-
tory. But if a black challenger is able to overcome white opposition and
win of¬ce, most white fears are not borne out. Black leadership may
lead to marginal changes to a few aspects of black well-being, but for
the vast majority of the white community, the world under black lead-
ers is strikingly similar to the world under white leaders (Smith 1996;
Singh 1998; Mladenka 1989, 1991; Eisinger 1982; Tate 2003; Browning,
Marshall, and Tabb 1997; Marable 1992). Once black of¬ceholders have
the opportunity to prove that black leadership generally does not harm
white interests, uncertainty should fade, whites™ views of blacks and black
leadership should improve, and more whites should be willing to consider
voting for black candidates.1 Black leadership therefore serves an impor-
tant although dif¬cult to observe informational role.

White Uncertainty and Fear in Black Challenger Elections
To understand the important informational role black leadership plays
for the white community, one must ¬rst understand why white Americans
fear black leadership in the ¬rst place. When blacks run for of¬ces they
have never held, most whites do not know what to expect. Normally,
voters have less information about a challenger than they do about an

1 My information model is akin in some ways to the interracial contact hypothesis developed
by Allport (1954) and con¬rmed by others (Jackman and Crane 1986; Sigelman and
Welch 1993). The interracial contact theory maintains that social contact or friendship
with minorities of the same socioeconomic status should have a positive effect because
it places individuals from two groups in a cooperative setting where their similarities
become evident. In this way, social contact and friendship differ fundamentally from most
interracial interactions, which are often competitive in nature. I believe that the act of
experiencing black leadership may represent a different type of interracial contact that
also generally leads to a positive learning experience.
Changing White Attitudes

incumbent. When that challenger is black, information is even sparser:
most whites have not lived under a black mayor, a black representative,
a black senator, or a black president. They simply do not know what the
consequences of black leadership are likely to be.2 Whites express concern
about racial integration, school busing, and the ¬‚ight of white businesses,
to name just a few of their fears. A truck driver in Chicago described his
concern about a black mayoral victory in this way: “I don™t know how
to say this but I am afraid [Harold Washington] 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 is going to try to push racial integration,
which is ¬ne as long as I don™t lose money on my house . . . because I just
can™t take the loss” (quoted in Coleman 1983a).3
Having little or no personal experience with black leadership, many
white voters rely on heuristics, or shortcuts, to try to gauge how black
leaders are likely to behave once elected (Conover and Feldman 1989;
Rahn 1993; McDermott 1997). When the candidates under consideration
are black, the chosen heuristic is usually race (Reeves 1997; Terkildsen
1993; Williams 1990). Given the fraught history of race in this coun-
try, persistent economic differences between white and black America,
ongoing racial con¬‚ict in many cities and states, and sharp disagreement
between whites and blacks over government policy, white voters have
some logical reasons to assume that black leaders will try to serve black
interests, however wrong that assumption usually turns out to be (Kinder
and Sanders 1996; Schuman et al. 1997).
Black challengers can and usually do try to counter the uncertainty
surrounding their candidacies by running “deracialized” or pro-white
campaigns, but white voters tend to ignore these candidates™ campaign
statements, which they perceive as having little credibility (Lupia and
McCubbins 1998). Moreover, black candidates™ efforts to deracialize their
campaigns are often overcome by their white opponents, who attempt to
garner white support by playing on white fears. The media also some-
times heighten uncertainty by continuously noting the racial nature of
black-white contests and reminding voters in certain contests that a black
victory would put an African American in control of a particular of¬ce

2 Underlying this account is the belief that uncertainty, rather than ambivalence, is at the
heart of white views toward black challengers (Alvarez and Brehm 2002).
3 As Bartels (1986) has shown, uncertainty in and of itself is likely to hurt a candidate™s
chances of winning an election, so the extraordinary uncertainty surrounding black chal-
lengers puts them at a severe disadvantage with white voters.
Black Leadership: The Possibilities 17

for the ¬rst time (Mendelberg 2001; Reeves 1997; Graber 1984). In the
end, the limited information that whites generally have, the in¬‚ammatory
campaigns that white opponents frequently run, and the racial heuristics
that whites rely on all tend to fuel negative projections about the impact
of a black victory. Heightened uncertainty in black challenger elections
leads most whites to believe they are facing an anti-white candidate and
to vote to prevent a black takeover.4

Why Does Black Incumbency Make a Difference?
Incumbency gives any of¬ceholder a critical advantage (King and Gelman
1991). Often, we think of that advantage in terms of an incumbent™s access
to resources or endorsements, but incumbency also plays a vital informa-
tional role. As Popkin has noted, “The incumbent is, to a certain extent,
a known commodity. In contrast, a challenger is often a great unknown”
(Popkin 1995: 33). As residents become more and more familiar with an
incumbent and his or her actions, their uncertainty and fear about what
he or she might do slowly fades away. Because uncertainty surrounding
black candidacy is usually much greater and misperceptions about the
consequences of black leadership are more widespread, this process is all
the more important for black representatives.
The key is that whites believe the information they get from their expe-
riences with black leadership is credible. It is not simply cheap talk (Lupia
and McCubins 1998, Alvarez 1997). Experiments in social psychology
have increasingly found that one of the few times whites really change
their attitudes about blacks is when they know that blacks are free to
choose their actions. Wilder, Simon, and Faith (1996), for example, main-
tain that whites™ stereotypes of blacks change only when whites witness
African Americans who act anti-stereotypically and when they view that
behavior as internally caused. In short, only when blacks can wield author-
ity are their actions likely to be seen by whites as truly informative. Thus,
black political leadership is especially important in the minds of white res-
idents because it marks one of the ¬rst times that blacks have authority

4 This pattern of white mobilization in the face of a new challenge from blacks has been
demonstrated in a range of political interactions between blacks and whites. Research has
shown that as black voter registration increases, whites also increase their registration
rates (Alt 1994; Loewen 1990). Racial bloc voting, already prevalent throughout the
country, increases in areas and districts where blacks become active voters (Mayer 1996;
Murray and Vedlitz 1978).
Changing White Attitudes

to enact policies or make changes that could harm the white community.
Once black representatives have had a chance to govern, and thus an
opportunity to assist the black community at the expense of the white
community, whites obtain important information about the interests and
preferences of black representatives. If black of¬ceholders are perceived
to have voluntarily pursued an agenda that helped both the black and
white communities, then whites who feared a black takeover may reeval-
uate their opinions of black leadership: when white residents do not lose
their jobs, when blacks do not move into white neighborhoods in large
numbers, and when crime does not proliferate under a black mayor, white
voters learn that they have less to fear from black leadership than they
originally thought.5 If they choose to, these voters can begin to base their
assessments of black leadership on black incumbents™ track records rather
than on stereotypes, exaggerated fears, or the incendiary predictions of
white candidates, leading to a more limited focus on racial considerations
in subsequent biracial elections.
If this information model is valid, we should see a distinct pattern
in elections in which black and white candidates oppose each other. In
black challenger elections, because most whites™ decisions about whether
to support the black candidate will be based on racial fear, few should
choose to support the black candidate. In black incumbent elections, on
the other hand, racial fears should play a diminished role, and a greater
number of white voters should cast ballots based on the track record of the
incumbent and the speci¬cs of the campaign. As more blacks are elected to
leadership positions over the years, uncertainty regarding black leadership
should decline in American society as a whole, and whites should become
increasingly inclined to consider supporting black challengers.
An important assumption behind this model is that whites have infor-
mation about black leadership. This may be a dif¬cult assumption for
some to accept. As I mentioned in the introduction, existing studies sug-
gest that the average American knows very little about politics (Ber-
elson, Lazarsfeld, and McPhee 1954; Bennett 1995; Campbell et al. 1960;

5 Survey experiments have shown that when given information about blacks that clearly
contradicts stereotypes, whites™ political views do change (Pef¬‚ey, Hurwitz, and Sniderman
1997). Pettigrew (1976) and Eisinger (1980) were among the ¬rst to notice the impact of
black incumbency on white fears. More recently, Swain has argued that information may
be a key variable for black candidates: “It is instructive that the black candidates who
have been most successful in winning white support typically have provided the voters
with plenty of information about themselves” (1995: 209).
Black Leadership: The Possibilities 19

Converse 1964; Delli Carpini and Keeter 1996).6 There is ample evidence,
however, that information levels vary across contexts. Numerous studies
show that the public is especially well informed about the select issues
that they care about (Hutchings 2003; Iyengar 1990). Thus, if the elec-
tion of a black candidate to of¬ce is something that many white Americans
perceive as an especially important or especially threatening event, these
individuals are likely to have or to acquire information on the subject. It is
clear, furthermore, that many white Americans do care who their mayor
is. As Thomas Pettigrew (1976) has noted, “Running for captain of the
ship” is the ultimate test of how averse whites are to black control. Even
in cities with weak mayors, the mayor is seen as the symbolic leader of
the city and the main person responsible for local conditions (Coleman
and Sussman 1978). If the city falls apart, the mayor is likely to be the
¬rst one blamed (Holli and Green 1989; Stokes 1993). And as we will see,
the typical white-black contest sparks extraordinary attention from the
electorate. For white Americans who care about race, then, the election
of a black mayor is an important step worth paying attention to.
Moreover, white voters do not need a lot of information to assess
black leadership.7 They need only know two things: 1) that an African
American is in of¬ce, and 2) that their own well-being or the well-being
of friends has not been negatively affected.8 More sophisticated voters
may learn speci¬cs about the policies and actions of black leaders and
acquire data on the economic and social well-being of the black and white
communities at different points in time, but such detailed information is
far from necessary for an individual white resident to update his or her

6 Page and Shapiro remark that “it is undeniable that most Americans are, at best, fuzzy
about the details of government structure and policy” (1992: 13). The incoherence of
individual public opinion may, however, be overstated by these studies. Other accounts
have suggested that much of the instability of individual opinion is due to measurement
error (Achen 1975) and real ambivalence about issues (Hochschild 1981; Alvarez and
Brehm 2002).
7 An emerging trend in the recent literature on information is to argue that individuals do
not need a lot of information to make reasonable choices. Individuals can use a variety
of “cognitive heuristics” or shortcuts to simplify the political environment and help make
rational decisions (Tversky and Kahneman 1974; Simon 1945; Lupia 1994; Popkin 1991;
Iyengar 1990). Some have even found that because political choices are often simple, those
with limited cognitive ability tend to make decisions in a very similar fashion to those with
the highest levels of cognitive ability (Rahn et al. 1990).
8 Such information is not usually dif¬cult to obtain: as I have mentioned, media coverage of
white-black political transitions is usually extensive and often highlights the signi¬cance
of new black leadership (Graber 1984).
Changing White Attitudes

view of black leadership. In other words, the information hurdles to white
learning are not high.
Fortunately, for the information model, the available evidence suggests
that most white residents do have both critical pieces of information.
First, white residents are very likely to know about the existence of black
leadership. Surveys indicate that most residents generally know who
their mayor is and that, when the mayor is black, almost all residents
are aware of the race of the mayor. Cole (1976), in particular, in a study
of twenty-¬ve cities in New Jersey found that roughly 80 percent of city
residents could identify their mayor. Even more importantly, having a
black mayor increased knowledge and interest. In cities with black may-
ors, roughly 90 percent of respondents could identify the mayor. Other
polls in other states and cities have partially corroborated these results.9
Second, there is evidence to suggest that most white Americans do get
enough information to evaluate black leadership. As we will see in the
two case studies, and at various points throughout the book, when asked,
the vast majority of white residents are willing to evaluate local economic
conditions and the performance of their local (black) leaders. Moreover,
these evaluations have a substantial impact on voting preferences in
subsequent elections.10 This is evident both in the case studies presented
later in this book and in several recent studies of mayoral approval, which
show a close link between evaluations of a range of speci¬c city services
and mayoral approval under black and white mayors (Stein, Ulbig, and
Post 2005; Howell and Perry 2004; Howell and McLean 2001). In short,
most white residents of cities with black mayors seem to have enough

9 A May 2000 Kaiser poll found, for example, that, nationwide, 58 of every 100 respon-
dents knew their mayor. When that mayor was a big city black mayor, knowledge levels
seemed to be even higher. A May 1992 Gallup poll indicated that 92% of respondents
nationwide had heard of and had an opinion on Tom Bradley. The ¬gure for David Dink-
ins was 72%. The same survey indicated that even Richard Arrington was known by some
46% of the national population. Other surveys likewise demonstrate that mayors tend
to be more well known than Congressmen or state representatives (Lewis, Taylor, and
Kleppner 1997).
10 We also know that in other contexts, e.g., national politics, there is ample research
indicating that individual Americans can accurately gauge trends in economic circum-
stances. As Conover and her colleagues note, “the public is remarkably accurate in their
assessments” of national economic trends (Conover, Feldman, and Knight 1986: 574).
Moreover, data from a variety of elections suggest that individual voters can and do use
these assessments to help determine their vote choice (Nadeau and Lewis-Beck 2001;
Kiewiet 1983). Incumbents who preside over expanding economies are likely to garner
more votes than incumbents whose tenure coincides with dips in real income or increases
in unemployment (Erikson 1989; Tufte 1978).
Black Leadership: The Possibilities 21

information to at least consider changing their views of blacks and black

enduring racial stereotypes
Believing that white residents will respond rationally to new information
and will be open to change on matters of race may very well be na¨ve; ±
the availability of credible, positive information about black leadership
does not guarantee that whites will assimilate it and change their views.12
Traditional accounts, in fact, seldom suggest that the white population
is receptive to change where race is concerned (Key 1949; Allport 1954;
Tajfel 1981; Dovidio and Gaertner 1986; Kinder 1986; Jackman 1977).
Instead, scholars are more apt to argue that because the bulk of the white
community is guided by racial animosity and racial prejudice, there is little
reason to suspect that white attitudes or behavior will change in response
to black leadership (Allport 1954; Hurwitz and Pef¬‚ey 1998). Prejudice,
“an emotional, rigid attitude” that is “irrationally based,” is likely to
be too deeply embedded to be easily discarded (Pettigrew 1972; Fazio
et al. 1995; Rothbart and John 1993). As Allport notes, “A prejudice,
unlike a simple misconception, is actively resistant to all evidence that
would unseat it” (1954: 9). Even if the words and actions of black incum-
bents do not ¬t whites™ racial stereotypes, a prejudice model predicts that
whites will use an array of tactics to try to maintain those stereotypes
and create cognitive consistency (D. Hamilton 1981). They will ignore
events that do not square with their views of blacks and discount con-
tradictory evidence as an exception to the rule (Kunda and Oleson 1997;

11 It is worth making one last comment about information levels: much of the recent liter-
ature on political information notes that aggregate opinion is much more rational and
coherent than individual opinion. By aggregating preferences, we can often cancel out
measurement error and individual mistakes to obtain views that are more responsive
to real-world events (Page and Shapiro 1992; Erickson and Wright 1989; Bowler and
Donovan 1998; but see Althaus 2003; Bartels 1996). As Hutchings notes, “Collective
opinion is often remarkably informed and in¬‚uential” (2003: 6). Thus, when we focus
on the aggregate white vote or average white views, we stand a good chance of seeing
opinion change that is directly related to changes in local conditions.
12 Of course, a critical assumption behind this information model is that white residents use
this new information and change their minds as a result of learning that black leadership
does not harm the white community. Sniderman and Piazza (1993) have been able to
demonstrate marked changes of opinion on matters of race when respondents are exposed
to new information and arguments, but ultimately the only way to test the assumption
that information leads to white learning is to look for changes in whites™ views and votes
before, during, and after the election of a black leader.
Changing White Attitudes

Rothbart and John 1993; Macrae, Hewstone, and Grif¬th 1993; Weber
and Crocker 1983).
But is this a realistic picture of white America today? Race relations
have been fundamentally transformed in the last century (Klinkner and
Smith 1999; Schuman et al. 1997; Thernstrom and Thernstrom 1997).
Expressions of biological racism have declined precipitously. As Kinder
and Sanders comment, “Remarks, once thoroughly representative of a
particular time and place, are unimaginable today” (1996: 92). White
Americans offer almost unanimous support for the principle of racial
equality (Schuman et al. 1997). And there are those who believe that
whites are no longer fundamentally driven by racial animosity (Sniderman
and Carmines 1997; Thernstrom and Thernstrom 1997). Still, there is
ample evidence that racial stereotypes and racial prejudice continue to
play a central role in American politics. Surveys indicate that much of
the white community stereotypes blacks and believes that blacks as a
group are not as intelligent, not as hard working, more violent, and more
disagreeable than whites (Bobo and Johnson 2000; Lee 2000). Longi-
tudinal studies of stereotypes, in fact, ¬nd that whites™ views of blacks
have changed little in recent decades (Devine and Elliot 1995; McCona-
hay, Hardee, and Batts 1981; Schuman et al. 1997). Moreover, a range
of research indicates that these racial considerations play a critical role
in white political decision making. Racial considerations in¬‚uence many
of the policy choices whites make (Gilens 2001), strongly shape white
partisanship and white voting patterns (Mendelberg 2001; Carmines and
Stimson 1989; Pef¬‚ey, Hurwitz, and Sniderman 1997; Edsall and Edsall
1991), and, most importantly for the purposes of this study, affect white
willingness to support black candidates (Terkildsen 1993; Reeves 1997;
Colleau et al. 1990). In one study, white respondents who knew little
about candidates besides their race rated black candidates as worse than
white candidates on 19 out of 20 leadership and personality characteris-
tics, viewing black candidates as less trustworthy, less able to “get things
done,” and less intelligent (Williams 1990). And in many real-world con-
tests, such stereotypes clearly prevent whites from voting for black can-
didates (Kinder and Sears 1981; Pettigrew 1972; but see Citrin, Green,
and Sears 1990; Highton 2004). As one scholar put it, “Black political
aspirants cannot compete equally or effectively in electoral jurisdictions
comprised overwhelmingly of white voters because of the continued vigor
of racial prejudice and discrimination” (Reeves 1997: 9).
Thus, the prejudice model predicts that no matter how well the white
community does under black leadership, black incumbency will ultimately
Black Leadership: The Possibilities 23

have no effect on whites™ attitudes toward blacks or on whites™ behavior
toward black incumbents. Because racial prejudice is widespread, this
model argues, there is little reason to believe that black of¬ceholders can
change the way white Americans think about race or the way they vote
in black incumbent elections. In the end, no matter what black leaders do
or do not do in of¬ce, few whites are likely to be signi¬cantly affected by
their experiences with black leadership.

white backlash
Another view in the literature on racial politics proposes that white actions
and attitudes, rather than stemming primarily from prejudice and racial
animus, instead re¬‚ect concerns about racial hierarchies or social status
(Sidanius, Devereux, and Pratto 1991). According to this view, whites
may dislike blacks and hold a range of stereotypes about them, but what
they care most about is maintaining their hegemonic position in society.
As Sidanius and his co-authors note in their outline of social dominance
theory, there is “a very general and basic human desire to perceive one™s
group as superior to and possessing greater social status than the general-
ized other” (Sidanius, Pena, and Sawyer 2001: 380). To the extent that this
is true, having African Americans in positions of leadership poses a real
threat to whites. When blacks attain positions of power over the white
community, the pre-existing racial order is turned upside down, regardless
of what black leaders do in of¬ce. Since the stability of the racial order
is critical to the white community, there is every reason to believe that
whites will respond to black electoral victory with counter-mobilization.
As Piven and Cloward predicted in the late 1960s before the election of
the nation™s ¬rst big-city black mayors, “Negro control can only deepen
racial cleavages in the urban area” (1977: 17).
The backlash model of behavior is supported by the historical record.
Most infamously, Southern whites responded to the signi¬cant expansion
of black political representation during Reconstruction with massive resis-
tance, instituting a program of unprecedented violence, poll taxes, new
residency and registration requirements, and at-large elections (Parker
1990; Foner 1984; Holt 1979). In Louisiana, for example, in less than
one year, Democrats killed over one thousand people in their effort to
regain control of the political process (Kousser 1974). Over a thirty-year
time span, white Southerners virtually wiped out all of the gains made
by black voters: in 1872, there were 324 blacks elected to state legisla-
tures and Congress in the former Confederate states, but by 1900 only
Changing White Attitudes

5 black of¬cials were in power (Kousser 1992). More recently, the actions
of many whites during the civil rights movement reaf¬rmed the backlash
hypothesis, as white violence spiked in response to the passage of various
civil rights initiatives and numerous other encroachments on white privi-
lege (Stenner 1995). Olzak (1990, 1992) found that in subsequent years,
white racial violence was particularly pronounced during periods of uncer-
tainty or instability in race relations. Speci¬cally, rates of racial violence
rose in response to black migration to formerly white-dominated urban
areas, black entry into formerly white professions, and political chal-
lenges to white supremacy in the South. Green and his colleagues (Green,
Strolovitch, and Wong1998) have demonstrated that racial migration
was a key determinant of hate crimes against blacks in the late twentieth
century. As Klinkner and Smith (1999) have astutely noted, moreover,
progress toward racial equality has been uneven. Episodes of progress
and black gains have almost inevitably been followed by retrenchment.
The historical record is clear: at many points in American history, white
Americans have not welcomed blacks™ gains in the political arena. If past
patterns of white behavior in the face of black empowerment are any indi-
cation of how whites will respond to a modern-day transition from white
to black political leadership, we should expect black electoral victories to
be followed by heated white backlash.

variation across contexts, individuals, and leaders
Although these three theories are implicitly set up in opposition to one
another, no single model can explain the outcome of every biracial election
in American politics or the thinking of every white voter in those elections.
The intention of this study is merely to see which theory best accounts for
underlying trends or changes on the margin.
In addition, given that white responses to black leadership do vary
from one case to another, a second important aim of this study is to
understand why white responses differ across leaders, contexts, and indi-
viduals. Variations in the style and substance of black leadership, differ-
ent local demographic contexts, and differences in individual character-
istics may all affect how whites respond to their experiences under black
I use the logic of the information model to make predictions about
where we should see the most pronounced and positive changes in white
attitudes and behavior and where we should ¬nd ongoing white resistance
to black empowerment. It suggests that variation in white reactions to
Black Leadership: The Possibilities 25

black leadership should be directly related to variation in the level and
type of information that whites receive from black leaders. In particular,
three factors should shape the information environment and ultimately
determine how large segments of the white community react to black
leadership: the impact of particular black leaders, individual attitudes, and
demographic context.

Learning Across Leaders
Since the information model argues that experience with black incum-
bency changes white views by demonstrating to whites that black leader-
ship does not appreciably hurt white interests, white reactions should be
dependent, at least in part, on the words and actions of black leaders and
the ongoing well-being of the local white community. In particular, efforts
on behalf of the black incumbent to redistribute substantial resources from
the white community to the black community may increase white fears
about the likely long-term consequences of black leadership. Similarly,
regardless of the stated intentions and policy actions of black leaders,
economic downturns that affect whites should eliminate gains in white
support. Cities or districts where housing prices plummet or crime rates
soar “ two primary white fears “ should see little or no change in white
willingness to support black candidates.
At the same time, as I have mentioned, the hurdle for black leadership
is not high: black leaders need only prove to be better than the exagger-
ated fears of many white residents. Coleman Young, for example, was
arguably one of the most radical black mayors of his generation. He vig-
orously pursued af¬rmative action and often put the black community
¬rst, but by the end of his tenure in Detroit, only a tiny fraction of white
residents had lost their jobs or their homes because of his actions. Because
his policies were less radical than many whites expected, even his tenure
seemed to allay white fears (Eisinger 1980).13 In fact, even if black mayors
wanted to radically redistribute resources toward the black community,
the reality is that in most cases they probably could not. Even in cities
where they have a great deal of power, mayors generally cannot unilat-
erally redistribute wealth or reform the macroeconomy. State and federal
constraints, competition from other cities, and rival politicians all serve as

13 For many white Detroit residents, the de¬ning moment of Coleman Young™s tenure
was probably his efforts to prevent a race riot by intervening with black protesters in
1975 rather than his actions to enact liberal policies on af¬rmative action or police
Changing White Attitudes

more or less binding constraints on local political leaders (Eisinger 1983;
Peterson 1981). It seems clear, then, that a black incumbent can serve
the black community in important ways while still allaying white fears.
In the majority of cases, there should be at least some reason for many
whites who live under a black mayor to learn that their fears about black
leadership are not warranted.14
In the end, while most cases of black leadership should lead to greater
acceptance by whites, the extent of that change in any given case will be
contingent both on local conditions and on black leaders themselves. As
such, it will be important to consider local policy initiatives, campaign
rhetoric, and changes in the social or economic well-being of the white
community when assessing white reactions to black leadership.

Learning Across Individuals
The political views and biases of individual white people also factor into
the process of white learning. The same information can be interpreted
in different ways by different individuals. In this study, I examine par-
tisanship as a mitigating factor on individuals™ views. White Democrats
often largely agree with the political views of liberal black challengers,
and thus they are the voters whose views and votes are most likely to be
transformed by information garnered from black leadership. With little
reason other than racial fears to oppose liberal black challengers, white
Democrat voters may be especially sensitive to information that reduces
those fears. Conservative white Republicans, by contrast, may have mul-
tiple reasons, in addition to race, for opposing liberal black candidates.
Also, given what many argue is a clear link between conservatism and
anti-black attitudes (Kinder and Sanders 1996; Kinder and Sears 1981),
we might predict that white Republicans will be more resistant to the
information that black representation provides.
Individual characteristics other than partisanship may also in¬‚uence
whites™ reactions to black incumbents. Those who know more about the
policies and actions of black leaders and their effects on local conditions
might more readily learn from and accept black leadership. Individuals
who are educated and well informed, however, are also less likely to hold

14 At the same time, there are likely to be some exceptional cases in which the informa-
tion provided by a black incumbent con¬rms white fears. Marion Barry™s tenure in
Washington, DC, for example, probably did little to improve whites™ views of black
Black Leadership: The Possibilities 27

exaggerated expectations about black leadership in the ¬rst place. Thus, it
is not clear whether their behavior will change more or less than the behav-
ior of those who are relatively uninformed. One could also argue that
learning under black incumbents should be related to an individual™s ini-
tial level of racial prejudice “ that the least prejudiced whites will respond
most openly to new information about blacks. But again, because the least
prejudiced individuals are among those least likely to have opposed black
leadership initially, it is not clear whether we should really expect more
change among this group, and, in practice, it is extremely dif¬cult to assess
change in either group. It is almost impossible to evaluate change among
the least prejudiced without panel data, because without it we do not
know if those individuals who are least prejudiced before the election of a
black leader are the same individuals who are least prejudiced afterward.
Likewise, given that surveys rarely include questions about respondents™
knowledge of local politics, we cannot know who is more informed about
local events under black leaders.15

Learning Across Cities
According to the information model, in order for whites to change their
minds about blacks and black leadership, they need credible information.
They must be able to observe the power of a black leader and connect that
power to local conditions. In other words, whites must believe that blacks
have some measure of control over local events. If, for example, whites
are unsure whether the continued well-being of the white community is
due to the newly elected black mayor or to an obstructionist white city
council, they gain little information from the tenure of their mayor, and
their views and votes are unlikely to change.
The degree of control exercised by a black incumbent should vary sys-
tematically with the racial demographics of the city in which that incum-
bent holds of¬ce. I will explain how black control and hence informational
effects should vary predictably across three types of cities: minority white
cities, majority white cities, and racially balanced cities.

15 One could substitute educational attainment or knowledge of national politics as proxy
for information about local racial politics, but it is not clear how closely either measure
is associated with local knowledge. I did attempt to see if changes in attitudes were more
pronounced among whites with more education or higher levels of political knowledge.
The analysis, which was similar in form to the analysis of partisan differences that is
presented in Chapter 4, revealed no clear or consistent patterns.
Changing White Attitudes

The most credible information regarding the interest and intentions
of black leaders is likely to be found in white minority cities, such as
Oakland and Newark, and thus it is in such cities that we should ¬nd
the greatest amount of change in white attitudes and voting behavior.
In white minority cities, the perceived threat from black leaders should
be extremely high because whites are outnumbered and a black mayoral
victory is likely to signal a real transfer of political power. If elected, a
black mayor would likely have the support of a large black constituency
as well as the support of other black elected of¬cials. Blacks might not
have control of the city™s economic and business interests, but they would
very likely have some control over most of the city™s major political of¬ces
(Grofman and Davidson 1994). If a black leader wanted to pursue radical
changes in policy, that leader would likely have the means to do so and
white voters would know this. The possibility of such radical change and
the fear that this evokes should convince most whites to vote against any
black challenger. After a black challenger wins the mayoralty, the black
community seemingly gains the power to advance its agenda. Although
this initially increases many whites™ fears of radical social and economic
change, it also allows black leaders to provide clear evidence of their policy
preferences and true intentions. When blacks enact agendas that do not
hurt whites, whites™ fears should abate. Thus, it is in white minority cities
more than anywhere else that white voters can draw lessons from black
In contrast, black leadership in majority white cities, such as Min-
neapolis and Seattle, should provide less information about the under-
lying preferences of black leaders, because black of¬ceholders in these
cities have little opportunity to take over the policy-making process, and
whites are aware of that fact. The prospect of a black mayor™s election
should spark little fear in majority white cities, because whether or not
the black challenger wins the mayoralty, whites will continue to dom-
inate the political landscape, and any successful politician will have to
cater to white interests. Since the threat most whites perceive from such
a candidate is low, many liberal whites should ignore race and vote for
the black challenger based on his or her nonracial quali¬cations. This
seems like good news for the black community, but because whites con-
trol the local political arena even after a black mayor enters of¬ce, the
information white voters receive about the black leader™s true preferences
is clouded. They cannot know whether the actions of the black incum-
bent are designed to serve black interests or simply to placate the larger
white community. The absence of any real change in white well-being
Black Leadership: The Possibilities 29

under a black mayor may convince some white residents that they have
little to fear from this black leader, but other white voters will dismiss
the mayor™s record on the ground that whites remained largely in con-
trol of the city. Such skeptical white voters will continue to fear the onset
of “real” black political control and to oppose black candidates, while
the white voters who supported the black challenger in the ¬rst place will
continue to support black leaders. In majority white cities, then, the infor-
mation model predicts that changes in white beliefs and behavior will be
Racially balanced cities, such as Chicago and New York, represent
a third distinct racial environment. What sets racially balanced cities
apart from other cities is their proximity to a racial tipping point. In
cities where blacks have about half of all registered voters, a signi¬cant
number of important political of¬ces, and other resources, the election
of a black mayor may be all it takes for the black community to gain
political control.16 This has important implications for the information
whites receive from black leadership, the level of white uncertainty regard-
ing future elections, and consequently the behavior of white elites. First,
proximity to a racial tipping point should spark intensi¬ed opposition
from white elites, who believe that losing control of local politics will
mean losing access to their jobs and livelihoods. Thus, in racially balanced
cities, black mayors should have a particularly dif¬cult time getting their
agendas enacted due to resistance from the of¬ceholding white elite. Each
action by a black mayor is likely to be followed by a counter action by
members of the white elite. As a result, few of the policies advocated by
a black mayor are likely to be enacted. Since black mayors are blocked
from achieving their goals, white voters receive little new information
from these mayors™ tenure in of¬ce, and learning is limited. Proximity
to the racial tipping point also leads to considerable uncertainty among
white residents about the future course of local politics. If blacks can
gain control of the local political arena at any point, every election could

. 1
( 7)