. 5
( 7)


focused on mobilizing the black community, tried to allay white fears
whenever possible.3 His campaign theme, “A Mayor for All Chicago,”
was part of an inclusive campaign that often explicitly told whites that
they would be included: “Our concern is to heal. Our concern is to bring
together. . . . I want to reach my hand in friendship to every living soul in
this city.” Washington also took more concrete steps to try to reduce white
fears. Jesse Jackson was “all but banished” during the general election

2 The importance of white fear was highlighted by the media and other political observers,
who generally concluded that the 1983 election was about white racial concerns rather
than about political reform, candidate quali¬cations, or any number of other potentially
relevant issues. The Washington Post called Chicago a city “engulfed in fear” (1983).
A review of public opinion polls taken during the 1983 contest simply concluded that
“Washington scared white people” (Rivlin 1992: 173). Another account in the Chicago
Tribune compared white fears in Chicago to the fears of white Southerners facing the civil
rights movement: “It is understandable why some segments of this city fear the kind of
change the election of Harold Washington as mayor signi¬es. It was the same fear felt by
the South during the civil rights struggle 20 years ago and in cities all over this country in
the years since” (quoted in Levinson 1983: 213). Another journalist tried to describe what
it was that whites feared: “Simply put, the fear in this and other white ethnic wards of the
city is that Washington™s election would cause a redistribution of power and resources.
White neighborhoods would suffer” (Peterson 1983b).
3 Washington™s campaign is best characterized as two campaigns, one focused on black
Chicago and the other focused on white Chicago. Each campaign had separate of¬ces
and different leadership. In front of white audiences, Washington pressed for racial unity,
¬scal conservatism, and reform. In front of black audiences, he preached an “It™s our turn”
Black Mayoral Leadership in Chicago 127

campaign, and Washington made sure he spent more time in white
Chicago than any of his opponents spent in black Chicago (Rivlin 1992).
He even announced the formation of a mostly white transition team in
the last weeks of the campaign to try to prove to white Chicagoans that
their interests would be considered.
Despite Washington™s efforts, racial fears appeared to dominate the
white vote.4 The aggregate vote total is the most obvious evidence of this
fear. In the Democratic primary, Washington received only 8 percent of the
white vote. In the general election only about 20 percent of white voters
supported him. In short, few whites in Chicago seemed willing to vote
for a black mayoralty. What makes this white opposition to Washington
more telling is that it broke voting patterns that had held for decades. To
vote against Washington in the general election, white Chicagoans had to
support the Republican nominee “ something they had been loathe to do
since the 1930s. In the decades before Washington entered the scene, the
Republican nominee for mayor had never garnered more than 5 percent
of the vote. But when Harold Washington ran on the Democratic Party
ticket, 79.3 percent of all white Democrats did the unthinkable and voted
Republican (Kleppner 1985). An “unchallenged bastion of Democratic
voting strength” had suddenly become a competitive bipartisan city.5 And
whites clearly did not ¬‚ock to Epton because he was a great candidate: he
was a total unknown prior to the campaign, had never won an election,
was regularly criticized for being a poor speaker, and was viewed as being
moody and temperamental, having twice spent time in a mental hospital.
By all accounts, if he had not been facing a black candidate seeking to
become the city™s ¬rst black mayor, Epton would have garnered next to
no support.6 What seemed to motivate white voters was race. A range of
analyses of exit polls, precinct returns, and pre-election surveys all found
that race was more critical than any other issue in the election (Baker
and Kleppner 1986; Day, Andreason, and Becker 1984; Kleppner 1985).7
White fears and racial concern were also re¬‚ected in remarkable white
voter turnout. A record 83 percent of eligible white adults registered for

4 In the end, the only thing that saved Washington was record black turnout (80.1% of
registered voters) and all but unanimous black support (Lewis, Taylor, and Kleppner
5 Moreover, the 1983 election did not mark a trend to more Republican voting. Since
Washington died, no Republican mayoral candidate has won more than 5% of the citywide
6 Overall, Bernard Epton won 49% of the vote.
7 My own analysis of an NBC/Associated Press primary election poll suggested that the
single most important determinant of the white vote was how important whites felt it was
that Harold Washington was black.
Changing White Attitudes

the general election, and of these registered voters a record 80 percent
turned out (Lewis, Taylor, and Kleppner 1997). The actions of white
Chicagoans when faced with the prospect of black leadership were not
all that different from those of whites in other cities. What set Chicago
apart is what happened after Washington entered of¬ce.

what information did washington™s tenure provide?
If whites in Chicago were as concerned about the consequences of black
leadership as were whites elsewhere, why didn™t experience with black
leadership reduce these concerns? The simple answer is a lack of informa-
tion. By all accounts, Washington was never given a chance to govern, and,
as a result, could not prove that black leadership would not hurt white
interests. The efforts of white Chicagoans at various levels inside and out-
side of city government successfully blocked almost all of Washington™s
The most visible and most effective resistance came from white mem-
bers of the city council. Even before Washington™s inauguration, his white
foes on the city council had hatched a plan to usurp power. During the
¬rst session of council, the anti-Washington coalition changed the elec-
toral rules of the council and took charge of all of the city council com-
mittees. The infamous council wars were on. Chicagoans quickly grew
accustomed to the rhythm of city politics under this new administration:
“Washington or one of his council allies, “ the Washington 21 “ introduced
an ordinance. It failed twenty-nine to twenty-one. The Vrdolyak 29 intro-
duced an alternative, which passed twenty-nine to twenty-one. Washing-
ton vetoed it, and there the matter remained deadlocked” (Rivlin 1992:
233). This pattern continued for almost three and a half years. Over that
period, almost every single ordinance that Washington™s administration
introduced was voted down.
In the ¬rst year, a vital downtown development faltered, a proposed
new public library failed to move forward, a historic Chicago theater
rejuvenation project could not get off the ground, and a key piece of city
real estate, the Navy Pier, remained undeveloped. Even O™Hare airport
was threatened with a shut-down in 1984. Washington™s ¬rst two budgets
were rejected by the Vrdolyak 29. What™s more, Washington could not
get much done with city government, because he was not able to gain
in¬‚uence over the bureaucracy. His appointments to various positions in
city government were consistently held up by white opposition on the city
council. Over eighty key Washington appointments were held hostage.
Black Mayoral Leadership in Chicago 129

The appointments languished “ two hundred days later, four hundred
days later, six hundred days later (Rivlin 1992: 263). Thus, from day one
of Washington™s term, his administration was essentially cut off at the
knees. For every action by the Washington™s administration, there was an
equal and opposite reaction. A headline in the weekly magazine Newsday
said it all: “Harold Washington: In Charge, But Not In Control” (Miller
The inability of Washington™s black-led coalition to pass its preferred
policies ultimately meant that black representation provided very little
information to white Chicago. Because Washington™s agenda was never
enacted, white residents in Chicago could neither reward him nor pun-
ish him for conditions in Chicago, and they could not know what would
happen if Washington and his black-led coalition were ever able to gain
control of the local political arena. Whites could have inferred something
from what Washington™s coalition wanted to pass. But even here, the mes-
sage was mixed, and thus the likely consequences of black control some-
what unclear. On the one hand, Washington pursued policies that were
not threatening. He took steps to try to balance the budget and cut the city
payroll. He also talked repeatedly about reform and being fair and open to
all parts of Chicago. But on the other hand, Washington took actions that
whites could easily have perceived as hostile. He tried to push af¬rmative
action and the redistribution of funds from downtown to the neighbor-
hoods. He also proposed record-breaking tax increases.8 He occasionally
made statements that whites could interpret as a threat to their ongoing
well-being. At one point, for example, he asserted, “Every group, when
it reaches a certain population percentage automatically takes over. They
don™t apologize. . . . they just move in and take over” (quoted in Graber
1984: 72). In short, Washington™s record was too limited, too inconsis-
tent, and had too many potentially threatening elements to convince most
white voters that they had little to fear from black leadership.
What whites could have easily learned over the course of Washington™s
tenure was that blacks were getting closer and closer to controlling the
local political arena. The last six months of Washington™s ¬rst term had
already tilted the precarious balance of control in favor of the African
American community. Special city council elections gave pro-Washington

8 Only on taxes was Washington successful. In 1987, he was able to pass major tax hikes. By
contrast, over the four-year period from 1983 to 1987, redistributive expenditures (social
services, education, housing, and community development) actually decreased by 0.3%
as a percentage of total government expenditures (Bureau of the Census 1964“2003).
Changing White Attitudes

forces control of exactly half of the city council seats. With the council
vote divided 25:25 on most issues, Washington could regularly determine
the outcome by casting the deciding vote.9 Moreover, although many of
Washington™s efforts at af¬rmative action had been stymied, there was,
nevertheless, slow growth in the African American presence at all levels of
city government. Under Washington, the proportion of city government
workers who were black rose from 26 to 30 percent, and the proportion
of black policy makers increased from 23 to 33 percent. Blacks still did
not hold enough positions to unilaterally control the operations of city
government, but with some help from liberal whites and other minorities,
they were very close to attaining a dominant governing coalition. Finally,
whites were losing ground in the electorate as well. In 1980, whites made
up over half of Chicago™s voting-age population (54 percent), but that
number was rapidly declining (down to 47 percent in 1987). The pen-
dulum was swinging, and it was clear that another Washington victory,
coupled with electoral success on the council and other arenas of local
politics, could push blacks over the top.

ongoing white concerns
Given that Washington™s tenure offered little information about black
leadership to white residents, and given that blacks were that much closer
to gaining control of the local political arena, one would expect whites
who had feared black leadership in the ¬rst place to continue to be anx-
ious about black representation. This is exactly what a review of events
in Chicago reveals. An array of evidence indicates that many whites in
Chicago continued to be concerned about the possibility that blacks would
take over.
White concerns were re¬‚ected most directly in public opinion polls
throughout Washington™s tenure. Early in his term, polls showed that
three-quarters of whites thought Washington was “out to get them”
(Rivlin 1992: 243). When asked their opinions in later polls, white
Chicagoans continued to state their fear of Washington and black leader-
ship. Washington™s own pollster continued to ¬nd that whites perceived
Washington as “hostile and threatening” (Rivlin 1992: 240). Donn Bailey,
a black sociology professor at Northwestern University, described the
white response to Washington™s campaign for reelection: “There™s a deep

9 With that power, Washington™s coalition managed to pass signi¬cant legislation, raising
taxes and expanding tenants™ rights.
Black Mayoral Leadership in Chicago 131

fear among the white people. I can understand it. They think we™re going
to treat them the way they™ve treated us” (quoted in Rivlin 1992). These
fears often spurred whites into action. Groups such as the Save Our Neigh-
borhoods, Save Our City coalition, an organization that fought black
encroachment into white neighborhoods, rapidly expanded their mem-
berships under Washington.
Not everyone openly expressed these fears, and, by some accounts,
white fears had moderated somewhat by the time Washington ran for
reelection.10 One local pundit argued that “the city is calmer now, because
white Chicagoans have learned that a black mayor is not the end of the
world. Harold Washington has served four years and the Sears tower is
still standing” (Bosc 1987). Several white residents recounted to reporters
the positive lessons they had learned under four years of black leader-
ship in Chicago. One retired white security guard, for example, pointed
to the lack of racial tension and the provision of city services in white
neighborhoods: “Actually Washington has performed pretty good. As far
as him splitting the city in half, I don™t think so. I don™t see marches,
things like that. It™s peaceful in the city. Police and ¬re protection, garbage
pickup “ they™re as good as they were before he got elected. Who knows?
I may end up voting for [Washington]” (Camper 1987). But in the end, it
seemed as if Washington™s tenure was not enough to convince the security
guard and others like him to actually support black leadership. Judging
by the aggregate vote, most white Chicagoans remained wary of blacks
and black leadership. In his reelection bid, Washington won few new
white converts. In fact, after four years of black leadership, white support
marginally declined. Washington garnered only 14 percent of the white
vote in the primary and a similarly small 15 percent of the white vote in
the general election.
More telling than the aggregate white vote is an account of why indi-
vidual white Chicagoans continued to oppose Washington. Unlike in most
other cities with black incumbents, whites in Chicago were often willing
to admit that race and racial fears still mattered. Almost three-quarters felt
that race would be a central factor for white voters (Neal 1986). More than
40 percent of whites admitted to pollsters that they would personally con-
sider race when voting in the upcoming election (Davidson 1987). Even
more important, unlike the white vote in reelection bids in other cities, the

10 White turnout did decline, suggesting that at least some white concerns had diminished.
Turnout of registered white voters dropped from a record high of 67% in 1983 to 59%
in 1987 (Lewis, Taylor, and Kleppner 1997).
Changing White Attitudes

white vote in Washington™s reelection bid did not seem to be closely tied to
Washington™s record. Despite the fact that more than half of white voters
(55 percent) thought that race relations in the city had actually improved
or were not getting worse under Washington, and the fact that half of
white voters believed that their own quality of life had either improved or
stayed about the same under black leadership, Washington garnered only
15 percent of the white vote. It was also impossible to point to party or
political ideology as the reason for the high level of ongoing white opposi-
tion to black leadership. Once again in 1987, almost 80 percent of white
Democrats voted against their own party to oppose Washington in the
general election. Finally, the white vote against Washington in 1987 was
clearly not a vote in favor of a particularly attractive white candidate.
Over a third of whites who voted for Washington™s opponent in the gen-
eral election admitted that they were voting to “stop another candidate
from being elected.”
What appeared to matter most was not what had happened over the
previous four years “ a period that offered little information to whites “ but
what might happen in the future. The possibility that the city would fall
apart if blacks gained control continued to motivate white Chicagoans.
One exit poll showed that concerns about the future were so extreme
that almost a third of white Chicagoans were “seriously thinking about
moving out of Chicago.” A northwest side resident put it in the fol-
lowing way: “Our neighborhood is really changing. We wonder what
the neighborhood will be like” (Belsie 1987). These concerns about the
future clearly affected the vote. Of those white Chicagoans who saw “bad
times” coming, only 4 percent supported Washington in the general elec-
tion. By contrast, 30 percent who saw good times ahead were willing
to support Washington. Overall, the vote in Washington™s reelection bid
was more like a black challenger election than a black incumbent election.
Party allegiances meant little, candidates meant little, and nonracial issues
played a secondary role. Black incumbency had not provided enough
information to convince whites that they need not fear another black
The nature of the 1987 campaign also suggested that racial concerns
were still prevalent in the white community. Although the candidates™

11 The range of ¬gures for the 1987 election are based on analysis of three citywide polls:
an ABC News primary pre-election tracking poll, an NBC News primary exit poll, and
an ABC News general election exit poll. Unfortunately, given the limited and incomplete
set of questions in each individual survey, no multivariate analysis could be performed.
Black Mayoral Leadership in Chicago 133

campaigns that year were, by most accounts, more muted than the heated
and racially explicit campaigns of 1983, there were still signs that both
Washington and his white opponents were acting as if white fears still
mattered. Both of Washington™s main opponents, Jane Byrne in the pri-
mary and Ed Vrdolyak in the general election, tried to highlight the dire
straits whites would be in if blacks were able to gain control of the
city (Grimshaw 1992). Vrdolyak was the most direct: “If [Washington]
is reelected, the sad fact is there won™t be another chance for us. . . . his
dream, our nightmare” (Rivlin 1992: 361). Byrne was more subtle, but
the racial tone of her primary campaign was also evident. White cam-
paign workers were also quick to use the ongoing threat of black lead-
ership to their advantage. Elmer Filipini, the 30th ward precinct cap-
tain, proclaimed, “If Washington wins, in four years they™ll be 100,000
or more whites moving out of the city of Chicago” (Secter 1987). In
response, Washington™s campaign tried to address these fears by muting
Washington™s rhetoric. There was no “It™s our turn” statement this time.
Washington smiled more, hired more white bodyguards, and avoided
most racial policy questions. But in the end, his actions did not mat-
ter. White Chicago had not changed its mind about black leadership. The
same racially divided world that Washington had taken charge of in 1983
was still there in the ¬nal year of his ¬rst term. The campaigns, the nature
of the vote, and the fact that Washington got only 15 percent of the white
vote all strongly suggest that white fears remained prevalent. Four years
had gone by but the theme seemed not to have changed: ˜Who knows
what will happen if blacks gain control?™

chicago after washington
Although Washington won in 1987, he had little opportunity in his second
term to demonstrate to white residents what black control of the local
political arena would do to the white community. Half a year into his
second term, he died of a heart attack. The city council then appointed
Eugene Sawyer, an African American who was part of the old machine,
as acting mayor. But Sawyer served as a caretaker for only a little over
a year before losing in the primary in 1989 to Richard M. Daley, the
son of longtime Chicago Machine boss Richard J. Daley. Black mayoral
leadership was over almost before it had really begun “ certainly before
black leadership had been given a real chance to prove itself.
The result was that the fears and concerns that white Chicagoans
expressed in various ways during the Washington years seemed to persist
Changing White Attitudes

table 6.1 White Support for Black Mayoral Candidates
in Chicago

Whites Voting
Election/Year Black Candidate Black (%)
Primary 1983 Harold Washington 8
General 1983 Harold Washington 20
Primary 1987 Harold Washington 14
General 1987 Harold Washington 15
Primary 1989 Eugene Sawyer 8
General 1989 Timothy Evans 6
Primary 1991 Danny Davis 7
General 1991 Eugene Pincham 2
Primary 1995 Joseph Gardner 5
General 1995 Roland Burris 6
Primary 1999 Bobby Rush 4
Sources: Lewis, Taylor, and Kleppner (1997); Alkalimat (1986); Rivlin
(1992); and author™s analysis of 1999 Chicago Tribune election poll.

in subsequent years. As illustrated by Table 6.1, over two decades after
the end of Washington™s tenure, the pattern of extremely racialized voting
continues.12 Although more recent contests have been void of the explic-
itly racialized campaigns that marked the election in 1983, the white vote
in each subsequent contest suggests that whites in Chicago continue to
have real concerns about black leadership. In every election since 1987,
black candidates have lost against a nearly unanimous white bloc vote. In
the ¬rst post-Washington election, Eugene Sawyer, despite his ties to the
Democratic Machine, garnered only 8 percent of the white vote in his pri-
mary bid in 1989.13 Subsequent black challengers have fared even worse
with white voters. Despite winning the majority of the black vote, Tim

12 There are certainly exceptions to this pattern at different levels of of¬ce. The successful
Senate campaigns of Carol Moseley-Braun and Barak Obama are two of the most promi-
nent. However, even in these two cases, it is not clear how supportive white Chicagoans
were of black leadership. Despite almost unanimous black support, Obama actually
lost Cook County in the primary before winning a majority of the white vote in Cook
County against another black candidate in the general election. Braun did win 53% of
the statewide white vote but she lost much of that white support in her reelection bid
(Oden 1996).
13 The 1989 election provided further evidence that white racial concerns about black
leadership, and not the candidates or the campaign, continue to determine the white
vote. The election turned out to be the most racially polarized contest in Chicago history
despite the fact that the campaign was “the least vociferous” in decades and the fact that
Sawyer was a longtime machine supporter.
Black Mayoral Leadership in Chicago 135

Evans, Danny Davis, Eugene Pincham, Joseph Gardner, Roland Burris,
and Bobby Rush all were opposed by more than 90 percent of white vot-
ers in their bids for the mayoralty between 1989 and 1999.14 This pattern
of ongoing white opposition suggests that little has changed in Chicago
and that much of white Chicago stills fears the onset of black control.

alternate explanations for continued white opposition
The information model offers a plausible explanation for why white
Chicagoans continue to oppose black leadership, but it is by no means
the only factor that could have contributed to ongoing white resistance to
black leadership. There are lots of reasons why white voters might oppose
black candidates and lots of other things that might make Chicago dif-
ferent from other cities. In this section, I consider a range of basic factors
related to the local political arena during Washington™s tenure that might
have played an important or even primary role in sustaining white oppo-
sition to his candidacy. Logically, if Washington™s tenure coincided with a
downturn in the local economy, if Washington or his opponents played the
race card more than they had in his ¬rst contest, if Washington was able
to raise less money in his reelection bid, if political endorsements favored
his opponent more in his reelection bid, or if he faced a more quali¬ed
opponent, then Chicago might be different from other cities, not because
of the lack of information black leadership provided but instead because
local political conditions turned against the black candidate.
However, none of these accounts offers a ready explanation for con-
tinued white aversion to black leadership. There was no marked decline
in local economic conditions, there was no increase in the racial nature
of the campaigns, and Washington enjoyed the usual advantages of
incumbency “ money, endorsements, and relatively inexperienced oppo-
nents. However, even after going through this process, I cannot de¬nitively
point to information as the cause of white actions in Chicago. Although
the four factors that I discuss are the main factors usually cited as being
critical in urban elections, they, by no means, represent an exhaustive set of
potentially relevant conditions. Other less quanti¬able aspects of Harold
Washington the candidate or Chicago the city could have contributed to
the outcomes. Harold Washington was, in many ways, different from the

14 Black voters were nearly united in their support for most of these black candidates. In
all of the elections between 1989 and 1999, black candidates for mayor garnered at least
three-quarters of the black vote.
Changing White Attitudes

typical black candidate. His occasionally overtly racial campaigns and his
vow to reform the machine suggested that more was at stake in Chicago
than in other cities.15 The city of Chicago, itself, was, in some respects,
also not typical. Intense racial segregation and a long history of a powerful
and sometimes corrupt Democratic machine are just two of the features
that set Chicago apart. With only one case and only one pattern to be
explained “ no decrease in white opposition to black leadership “ it is
impossible to rule out all of these potentially important in¬‚uences. The
available evidence may point to information as the main cause of sus-
tained white opposition in Chicago but any and all conclusions regarding
the racial learning process in Chicago are necessarily tentative.
Of the many factors that could have limited Washington™s support, a
downturn in the local economy is perhaps the most likely suspect. Over
and over again, economic conditions have been a central factor governing
the vote in American elections (Fiorina 1981; Erikson 1989). However,
economic conditions in the city of Chicago did not deteriorate under
black leadership (Keiser 1997; Fasenfest 1989; Wievel 1989; Fremon
1988). Contrary to the fears of many white voters, under Washington per
capita income actually increased, and unemployment ¬gures saw marginal

15 Did Washington lose some white support because he continued to press for reform of city
government? Probably. In both 1983 and 1987, Washington called repeatedly for an end
to patronage, the opening of the local governmental process, and an equitable distribution
of city funds. Unlike most previous politicians in Chicago, Washington also refused to
compromise with his machine predecessors once in of¬ce. This stance certainly frightened
white politicians, who opposed Washington at every turn. And the actions of white
machine politicians likely helped spur white opposition to Washington. Among other
things, the actions of white elites certainly helped to highlight the precarious balance of
power in the local political arena and prevented black leadership from proving itself. But
it is also clear that a lot of white opposition to Washington preceded rather than followed
white elite actions. As soon as Washington won the primary in 1983 and it became clear
that he had a chance to take over the city, much of white Chicago mobilized into action.
As Kleppner and others have noted, “It was a spontaneous and enthusiastic outpouring
from the grass roots, a groundswell. At that point, most of the white committeemen
were probably still stunned by Washington™s nomination and . . . took the election of
any Democratic nominee for granted” (Kleppner 1985: 191). Moreover, in 1987, white
Chicagoans continued to oppose Washington even though white machine politicians
were ¬ghting each other to see who would challenge Washington. Despite the fact that
the machine could not produce one “great white hope,” white voters still chose to vote
against Harold Washington in 1987. By almost all accounts, white fears had little to do
with reform issues either. As the New York Times put it, “Say what they will, Democrats
defecting in such numbers in Chicago are not just concerned for the probity of the
Democratic machine or candidates. Many white Chicagoans simply fear . . . the idea of
a black mayor” (1983a). In short, much of the sustained white opposition would likely
have been there with or without the issue of reform.
Black Mayoral Leadership in Chicago 137

improvement.16 Equally important, there was no marked change in the
relative economic status of the black and white communities. Blacks did
get a few more jobs, more appointments, and more contracts from city
government, but the change was far from dramatic and did not greatly
affect the welfare of most of the black or white communities (Joyce 1994;
Fremon 1988). The major problems facing poor, black Chicago were as
bad in 1987 as they had been in 1983.17 In short, there was little in
Chicago™s economy under black leadership that should have generated the
extraordinary white opposition Washington faced in his bid for reelection.
The lack of change in the white vote also cannot be attributed to the
increased racialization of either Washington™s or his opponents™ cam-
paigns. In 1987, both Washington and his opponents toned down the
racial rhetoric in their campaigns. Washington was by no means a color-
blind politician in 1987, but most agreed that his second campaign was
muted compared to his ¬rst (Rivlin 1992; Holli and Green 1989; Miller
1989). Washington™s campaign manager, Jackie Grimshaw, described it
this way: “Our approach this time is more intellectual than emotional”
(quoted in Rivlin 1992: 366). There was no Task Force for Black Politi-
cal Empowerment, no talk about taking over, and Washington “generally
steered clear of any mention of race” (Rivlin 1992: 366). After a passive
performance in a debate with Vrdolyak, Washington complained, “My
own campaign did something to me that Vrdolyak could never do. They
cut my balls off” (quoted in Miller 1989: 303). Instead of focusing on
race, Washington campaigned largely on his record, highlighting his ¬scal
responsibility while continuing to talk about neighborhoods and reform.
Nor did Washington™s opponents play the race card in 1987 to the same
extent they had in 1983. Racial slurs largely dropped out of the campaign
(Holli and Green 1989). Several ostensibly nonracial issues actually got
some air time (Grimshaw 1992; Fremon 1988; Davidson 1987). Byrne,
the primary challenger, talked a lot about crime. Washington™s general
election foes also spent time and resources addressing the issues of taxes
and Chicago™s poor schools. Race and white fears may have still been on
everyone™s mind, but, with a couple of notable exceptions, these racial
concerns were not the topic of conversation. Whites who voted against

16 Between 1983 and 1987, per capita income grew 5.7% (Bureau of the Census, 1964“
2003). Immediately prior to Washington™s arrival, 11.7% of Chicago was unemployed.
Just before his ¬rst reelection bid, unemployment stood at 9.3% (Bureau of the Census).
17 In 1987, poverty stood at over 20%, and one in ¬ve Chicagoans had no health insurance
(Fremon 1988; Rivlin 1992). The Chicago Housing Authority was still in crisis, and
Chicago™s public school system was viewed as one of the worst in the nation.
Changing White Attitudes

Washington in 1987 did not do so, then, because they had been exposed
to an exceptionally racialized campaign.
Washington™s inability to attract more white voters when he ran as an
incumbent cannot be explained away by pointing to a lack of incumbent
resources. In 1987, he was able to greatly outspend his opponents. In the
general election, Washington raised $6.3 million, over twice as much as
his closest competitor, Ed Vrdolyak, who raised $2.3 million. In 1987,
Washington also managed to get all but a few of the city™s major endorse-
ments, including ¬rm support from the city™s two major newspapers. He
was of¬cially endorsed by the Cook County Democratic Party. And, for
the ¬rst time, in 1987 many prominent local white Democrats actually
campaigned at his side, while other in¬‚uential Democrats chose to sit on
the sidelines rather than actively campaign for Washington™s opponent,
as they had done in 1983. There was no clear trend in candidate quality
and overall little link between the white vote and candidate quality across
Washington™s four primary and general elections. Whether Washington
faced the incumbent mayor, the assumed heir to the Daley machine, a
relatively unknown and inexperienced candidate, an ex-mayor trying to
make a comeback, or three white opponents (the most successful of whom
began with negative poll ratings unmatched in city history), his white sup-
port remained more or less constant (Holli and Green 1989).18 It appeared
that “anybody but Harold” was the choice of almost 90 percent of white
All told, Washington was a powerful incumbent who under normal
circumstances would have substantially increased his share of the white
vote. But in 1987, blacks were on the verge of real power, and nothing
about Washington™s track record, his opponents, or his campaigns had
demonstrated to white voters that their fears were unwarranted. After
four years of stalemate, fears of what would happen if blacks gained
control of Chicago politics appeared to be a critical force behind white
aversion to Washington and black leadership.

18 The fact that Washington could garner no more than 15% of the white vote in the
general election in 1987 despite facing three white opponents whose campaigns were
largely aimed at ousting each other and who, with the exception of Vrdolyak, whose
“disapproval” rating stood at 64%, had little name recognition and limited political
experience was perhaps the strongest sign that candidate quality was largely irrelevant
in these contests (Holli and Green 1989). As confusing and as unappealing as the white
candidates were, their shortcomings did not prevent white voters from uniting to serve
one electoral goal: the defeat of Harold Washington.
Black Mayoral Leadership in Chicago 139

Ultimately we cannot know for certain that a lack of information about
black leadership played the central role in sustaining white opposition to
black candidates in the city. There are just too many potentially relevant
factors and not enough data points to rule them all out. However, if there
is a lesson from Chicago, it seems to point to the importance of informa-
tion. Without credible information about how black leadership affects the
white community, little change seems to occur in the arena of local racial
politics. If black leadership is prevented from enacting its agenda, whites
get less information. As a result, fears about black leadership and highly
racialized voting patterns are likely to persist. Ironically, white racial fears
prevent white residents from learning that their fears are largely unwar-
ranted since the closer blacks get to gaining control and being able to
prove themselves to the white community, the more motivated whites
are to prevent that takeover. This conclusion does not bode well for the
future of race relations in Chicago. Today, whites represent only 42 per-
cent of the city™s population, and the balance of power between whites
and blacks remains precarious.19 Whites are very unlikely to want to cede
control at any point in the near future, and thus one could predict that
the black-white divide in Chicago will endure.
The future in other racially balanced cities seems less clear. The data
that I have presented in this book are not particularly encouraging. Across
all of the racially balanced cities that I examined in Chapter 4, white sup-
port for black candidates actually declined when those candidates ran for
the ¬rst time as incumbents. And even today, mayoral politics in cities
such as Philadelphia and New York can hardly be viewed as racially har-
monious. However, some racially balanced cities have shown signs of
improved race relations. In Houston, for example, the last two mayoral
elections have ended with multiracial coalitions supporting the winner.
In other racially balanced cities, black mayors have been given more of a
chance to enact their agendas. Black mayors in Dallas and Hartford, for
example, were able to put forward programs without incurring massive
white counter-mobilization. Whites may thus be getting important infor-
mation from black leadership in these cities. In addition, immigration is
altering the basic racial dynamics of mayoral politics in many racially
balanced cities. The growth of the Latino and Asian American popula-
tions in cities like San Francisco and Houston has certainly complicated

19 In large part, whites have been able to retain control by garnering support from the
growing Latino population.
Changing White Attitudes

the electoral equation, as both black- and white-led coalitions vie for the
support of these other minority groups. Thus, the future of race relations
in these cities depends upon a range of factors as diverse as racial learning
and immigration and there is at least some chance that white aversion to
minority leadership will diminish.

Other Cases Where Information Could Matter

This book is ostensibly about race “ about how black mayors affect the
views and votes of their white constituents. But the story presented in the
preceding chapters is also very much about information and how credi-
ble information can fundamentally alter individual views and intergroup
dynamics. In this chapter, I consider the implications of information for
other types of political transitions. I begin this task by reviewing how
and why information matters in the interactions between black mayors
and their white constituents. I then assess a variety of other cases to see if
they could plausibly ¬t an information-based account of behavior. In each
case, I ask two questions. Is the case structured in a way that suggests that
information should matter? Does the evidence point to learning? The goal
in each case is not to demonstrate or prove learning. In fact, the evidence
that I put forward in each case will be extremely limited. I merely wish
to point to cases where the presence of learning is plausible enough to
warrant further investigation.
According to the information model, the key to learning is that expe-
rience with new representation provides credible information. In order
to provide this, new leaders must be perceived to have real control over
outcomes and policies. The more that new leaders have the power to
in¬‚uence the well-being of one community, the more credible is the infor-
mation the members of that community get from their experience with
the new leaders, and the more likely they are to learn. Power and credible
information are not, however, the only conditions necessary for learn-
ing to occur. In addition, for credible information to lead to a change
in views and votes, residents must have exaggerated expectations. They
must have unrealistic concerns that stem from a lack of information and
Changing White Attitudes

the use of inaccurate stereotypes. For new leaders to reduce these fears
and stereotypes, moreover, they must not use their new power to harm
another group.1 It is the decision not to enact harmful policies that is both
surprising and informative.
To explore whether the information model can be applied more
broadly, I examine three additional cases below: whites™ reactions to black
members of Congress, African Americans™ reactions to black mayors, and
whites™ reactions to Latino and Asian American elected of¬cials. I also
very brie¬‚y consider whether the information model might be applied
to a range of nonracial political transitions both inside and outside the
U.S. context. Each case ¬ts the information model to a greater or lesser

black incumbents in congress
If information really does structure white reactions to black leadership,
it should not only explain variation in white responses across cities but
also account for different white reactions across different types of political
of¬ces. The more that a particular political of¬ce can control local policies
and affect local conditions, the more white residents should learn from
that position, and the more positively they should respond to the presence
of a black incumbent in that of¬ce.
Incumbents in Congress have a lot of power “ often more than mayors,
whose powers are limited by city charters “ but there are two key differ-
ences that make it easier to connect black mayors than black members of
Congress to local conditions. First, incumbents in Congress do not have
the power to act unilaterally. Unlike a mayor, legislators must obtain the
support of a majority of their colleagues to enact policies. Second, since
members of Congress are in Washington and often deal with national
issues, they are less frequently blamed or credited for outcomes at the local
level. Mayors, by contrast, are viewed as being primarily responsible for

1 Another important assumption behind the information model is that residents can and do
acquire information about the impact of the new leadership. Given the relatively low levels
of political knowledge evident in the mass public (Campbell et al. 1960; Delli Carpini and
Keeter 1996), there is some doubt as to whether most individuals are likely to acquire
information about new leadership. However, as I noted in Chapter 1, residents do not
need a lot of information to learn from new leadership. They need only know that new
leadership exists and that their own well-being “ or the well-being of their community “
has not changed under these new leaders. If the conditions of the information model are
met and residents do have real concerns about new leadership, then there is reason to
believe that most residents will have the motivation to acquire this limited information.
Other Cases Where Information Could Matter 143

local policy and local economic conditions. Moreover, although Congress
can appropriate considerable funds for local projects, their actions are less
likely to have a direct, visible effect on local conditions. From education
to police protection to garbage collection, the implementation of most
local policies occurs under the control of local government. All of this
means that it is harder to tie a local outcome to the actions and interests
of an individual legislator.2 If any politician is to be blamed or credited, it
will probably be the mayor. Since it is easier to track the effects of black
mayors than it is to track the effects of black members of Congress, black
mayoralties should provide more credible information to white residents
than do black Congress members. As a result, we should expect fewer
positive changes in white behavior under incumbent black members of
Congress than under incumbent black mayors.
In undertaking this comparison, I cannot provide the same analysis of
Congress members as I did for mayors. The main problem is that most
black members of Congress win in majority black districts and many, if
not most, face only black challengers. With few biracial elections and
few cases with large white voting populations, it is dif¬cult to undertake
rigorous empirical analysis of the white vote in all but a few elections. I will
simply review the relevant results from a number of existing studies that
have examined white voting patterns in black incumbent congressional
elections and present my own data on incumbent reelection rates.
To begin to see if black incumbency at the congressional level affects
white votes, I collected data on the outcome of every black incumbent
reelection bid in Congress in the twentieth century. As with black mayors,
the ¬rst point to note about black incumbents in Congress is that they win
almost all of the time. In the last century, African American members of
Congress won reelection an impressive 97 percent of the time (302 out of
312 cases). This rate of success puts black incumbents in Congress roughly
on a par with white House members, who generally win reelection over
90 percent of the time (Stanley and Niemi 1992).
What is more important for our purposes is how black incumbents
do when facing white voters. Here, the results are also fairly clear. In
the thirty-three cases where a black incumbent went up for reelection
in a majority white district, the reelection rate was an equally impressive

2 Another important factor limiting the information provided by black legislators is that
they are invariably in the minority in the legislature. In the House, black members hold less
than 10% of the seats. No state has close to a majority of black legislators. Thus, whether
a policy passes or not is usually not dependent on the preferences of black legislators.
Changing White Attitudes

94 percent. These reelection rates cannot tell us if black candidates garner
more white support as incumbents, but they do at least tell us that there
is no widespread white backlash.
Importantly, the success of black incumbents stands in stark contrast to
the failure of black challengers. Although there is no complete record of
how many African American candidates have run for Congress, it is clear
that few have been successful in districts where white voters controlled the
outcome. Of the 6,667 House elections in white majority districts between
1966 and 1996, only 35, or 0.52 percent, were won by blacks (Canon
1999). And even when blacks won, it appears that most white voters
opposed them. In the bulk of studies that have assessed white support for
successful black challengers, the results suggest that only about a third of
white voters ultimately voted for the black challenger (Bullock and Dunn
1997; Voss and Lublin 2001).3 In other words, white voters appear to
be resistant to the prospect of black leadership at the congressional level,
and many try hard to prevent blacks from winning of¬ce. Once a black
representative or senator is in of¬ce, however, this resistance seems to
fade. Even in cases where white voters could oust black incumbents in
Congress, they almost never choose to do so.
A set of ¬ve reelection bids deserves special attention here. In the early
1990s, the Supreme Court struck down the districts of ¬ve black House
members and effectively forced then to run for reelection in newly redrawn
districts that no longer had black voting majorities. Most civil rights
activists and many academics felt the Supreme Court decisions would
spell doom for black representation in Congress (Reeves 1997). Laugh-
lin McDonald, the director of the southern regional of¬ce of the Amer-
ican Civil Liberties Union, predicted “a bleaching of Congress” (Sack
1996). The Reverend Jesse Jackson called the Supreme Court™s move “a
kind of ethnic cleansing” (Sack 1996).4 But to almost everyone™s sur-
prise, the 1996 congressional elections told a very different story. In that
year, all ¬ve black members of Congress who ran for reelection in newly

3 In addition, there are often signs that racial prejudice is a determining factor in the white
vote in these contests. For example, the white vote in Edward Brooke™s attempt to become
the ¬rst popularly elected African American senator was closely correlated with indices
of prejudice (Becker and Heaton 1967).
4 Within the black community, anger at the court™s actions was pronounced. An editorial
written by one of the black incumbents stated: “Five Supreme Court Justices have done to
[blacks] in Louisiana what no hooded Ku Klux Klan mobs were able to do in this decade “
remove a [black] from Congress. The federal district court created a district where David
Duke, a former Klan leader, will have a far better chance of election than Cleo Fields [the
black incumbent]” (Fields and Higginbotham 1996).
Other Cases Where Information Could Matter 145

redrawn minority black districts won their races “ and with substan-
tial white support. In perhaps the two most surprising cases, Cynthia
McKinney won the 11th District in Georgia with the support of an esti-
mated 31 percent of white voters, and Sandford Bishop carried approx-
imately 36 percent of the white vote to victory in Georgia™s 2nd District
(Bullock and Dunn 1997). Equally important, but perhaps less surpris-
ing, none of these ¬ve black incumbents lost an election in the ensuing
years. Given that before 1996 less than 1 percent of previous elections in
the nation™s majority white congressional districts had produced a black
victor, this string of African American victories is quite striking (Canon
1999). It suggests that incumbency is a powerful tool for black candidates
at the congressional level. As Cynthia McKinney put it, “My victory says
more about the power of incumbency than anything else” (quoted in Bul-
lock and Dunn 1999: 15). As challengers, black candidates almost never
win when they face white voters. As incumbents, they almost always win.
The key test of the information model is not simply whether black
incumbents win or whether they win with white support but rather if
they win with more white support than they had as challengers. Although
data that allow an assessment of this change in the white vote are limited,
one study has attempted to measure change in white support during and
after the transition from white to black leadership. Bullock and Dunn
(1999) examine changes in white support between challenger and incum-
bent elections for a handful of black candidates who ran in the South in
the 1990s. The results varied from candidate to candidate, and some of
the black incumbents in their sample did lose white support over time.
However, when all of their cases were looked at together, white support
did increase. On average, black candidates garnered about 3 percentage
points more white support when they ran as incumbents in 1996 then
when they ran as challengers in 1992. This is certainly not an overwhelm-
ing shift in the white vote, but it may be enough to suggest that some
learning was occurring.
Another way to assess the importance of information and racial learn-
ing is to see if racial considerations prevent at least some white vot-
ers from supporting black incumbents. In other words, is the level of
white support for black incumbents less than it would be for similarly
situated white incumbents? Here, the results are quite mixed. In per-
haps the most expansive study of black incumbent congressional reelec-
tion bids, Highton (2004) used exit poll data to ¬nd that, after cam-
paign spending, partisanship, the presidential vote, and demographics
had been controlled for, white voters were not less willing to support
Changing White Attitudes

black incumbents than white incumbents.5 By contrast, Gay (1999) and
Voss and Lublin (2001) employed ecological inference across a smaller
number of Southern House elections to ¬nd that whites were, on aver-
age, 8 to 10 percent less likely to support black incumbents than white
incumbents. If Highton™s exit poll reports can be believed, it appears
that whites™ racial concerns largely disappear by the time black mem-
bers of Congress run for reelection. If estimated election returns more
accurately re¬‚ect the white vote, however, there is evidence that when
blacks run as incumbents, these fears, although not overwhelming, are still
If the information model works at the congressional level, white voters
should be more and more willing to support black congressional candi-
dates as more blacks gain of¬ce around the country. Here, the data are
much more limited but nevertheless suggestive. In the last four decades,
there has been a slow but steady increase in the number of black mem-
bers of Congress who represent majority white districts. The number grew
from zero in 1960 to one in 1970, two in 1980, four in 1990, and six in
2000. Although the number of blacks representing white districts is still
very small, the six blacks who come from majority white districts repre-
sent a signi¬cant percentage of all black members of Congress: in 2000,
16 percent of all black members of Congress represented districts where
whites were the majority of voters. Many factors likely contributed to this
expansion, but the increased willingness of white voters to support black
challengers for Congress certainly seems to have helped to push blacks
closer to parity in the House.
Much more research needs to be done, but overall, the evidence seems
to suggest that incumbency at the congressional level has a real but limited
effect on white voters. Incumbency would appear to lead to modest gains
in white support for black candidates and may be leading to a diminished
role for race in the voting booth. The modest changes that we see under
black incumbents in Congress ¬t well with an information model of white
behavior. Given that blacks in Congress exercise little direct control over
local policy or local conditions and given that it is dif¬cult to gauge the
impact of black leadership at the congressional level, one would expect
experience under a black incumbent to in¬‚uence only a small number of

5 Highton (2004) examines about 18,000 white votes in 357 contested House elections (37
with black incumbents) in 1996 and 1998.
Other Cases Where Information Could Matter 147

african americans and black elected of¬cials
Another interesting potential application of the information model is the
case of African American voters. In the accounts of mayoral politics that
I have presented in this book, I have not given a great deal of attention to
black voters. But clearly, in these cities and elections, black voters played
a critical role. Without their support, few black challengers would ever
have won of¬ce. Obviously, the fate of black incumbents is intricately tied
to the decisions made by black voters. Do they learn from their experience
with black mayoral leadership as well? Just as it was for white residents,
prior to the election of black mayors, black residents had not seen black
leaders in positions of real authority over the white community. Indeed,
much of the available evidence suggests that many members of the black
community had very high expectations of black political leadership. One
account notes that “everyone was optimistic life was going to get bet-
ter” (Donze 1998). Another refers to “an almost revolutionary rise in
expectations” (Span 1983: 55). Despite the fact that black residents likely
had more informed and more nuanced views of the black community
than did whites, there appears to have been considerable room for learn-
ing among African Americans.6 Equally importantly, African American
residents lived through the same events and witnessed roughly the same
information. Thus, there is at least some reason to suspect that informa-
tion should matter for African Americans as well.
A brief review of elections in the cities highlighted in this book sug-
gests that experience with black leadership did in fact change African
Americans™ views and political behavior. In the same way that experi-
ence under black mayors taught white voters that they have little to fear,
experience under black leaders seems to have taught African Americans
that they could not expect black leaders to be able to redress deep-seated
racial inequalities. The early euphoria surrounding black electoral victo-
ries seems in many cases to have been replaced by more reasoned assess-
ments of black leadership. As one black resident of New York put it, “It™s
become clear that [Dinkins] has not produced dramatic changes and prob-
ably won™t” (Hicks 1993). Others were even more direct. In Los Angeles,
a black voter complained: “To put it bluntly, [Bradley] just hasn™t done a

6 Surprisingly, it is not clear whether African Americans always hold different views of
the black community from whites. Surveys of white and black Americans reveal that
African Americans tend to hold similar stereotypes about the black community and black
candidates (Schuman et al. 1997; Williams 1990).
Changing White Attitudes

hell of a lot for black folk” (Litwin 1981: 88).7 A retiree in Gary expressed
a similar view: “It™s been 28 years of black mayors and I don™t see what
they have done” (Kendall 1995). Though this evidence is anecdotal, it
suggests that time under black incumbents has taught some African Amer-
icans that black political representation cannot easily or quickly solve the
black community™s basic economic problems.8
A drop in turnout is perhaps another sign of black learning and dimin-
ished expectations regarding black leadership. Intense mobilization and
record black turnout in black challenger elections slowly gave way to
average or even record low turnout in black incumbent elections. As one
reporter in Newark put it: “They danced in the streets in 1970. They
walked to the polls in 1974, and they crawled to the polls in 1978”
(Oreskes 1981: 29). Across the twenty-six cities examined earlier, black
turnout fell from an average of 66 percent in black challenger campaigns
to 59 percent when the same candidates ¬rst ran for reelection a few years
later. In many subsequent elections, black turnout dropped even further.
In Detroit, for example, black turnout slowly declined during Coleman
Young™s tenure to the point where just under a third of registered black
voters were going to the polls in his later bids for reelection (Rich 1987).
And in a range of cities from Los Angeles to New Orleans to Memphis,
successive elections with black incumbents similarly saw lower and lower
¬gures for black participation (Jackson and Preston 1994; Wright 2000).
Although a number of factors likely contributed to declining black parti-
cipation, frustration with the pace of change under black incumbents may
very well have played an important role. As one account put it, “the real-
ity of governance generated less enthusiasm than its prospect” (Peterson
1994: 2).
Even more telling, perhaps, is the increasing willingness of some black
voters to favor white candidates over black candidates.9 While the trend

7 At times, this frustration has been extreme. One black councilwoman in Cleveland was
particularly negative: “The mayor doesn™t represent the black community. The mayor is
bought and sold by the power structure in this city” (Milbank 1993: A1).
8 Despite what appears to be widespread frustration about the pace of change, African
Americans still seem to garner psychic bene¬ts from black representation. Tate (2003)
and Bobo and Gilliam (1990) have both shown that black leadership leads to increased
trust and ef¬cacy among African Americans.
9 This trend was not apparent in the ¬rst few years of black leadership. Across the 26 cities
examined earlier, black voters were just as supportive of black mayoral incumbents (95%
of blacks supported them on average) when they ran for reelection for the ¬rst time as
they were of the same candidates when they ran as challengers (93% of blacks supported
them on average).
Other Cases Where Information Could Matter 149

is by no means universal, a number of largely black cities that had long-
term black leadership have begun to return whites to the mayor™s of¬ce.
In 1995, for example, Gary, a city whose population is nearly 90 percent
black, elected a white man, Scott King, to the mayoralty. King captured
83 percent of the citywide vote in the general election, and in subsequent
electoral bids he handily won reelection. Similarly, in 1998, after twenty-
one years of black mayoral leadership, Jerry Brown took over as mayor of
Oakland, a city that is only 24 percent white. Baltimore, another majority
black city, also opted to elect a white candidate over a black candidate
in 1999 following over a decade of black mayoral leadership: Martin
O™Malley garnered one in three black votes on his way to victory over
two black candidates (Shields and Penn 1999). All told, almost half of the
large majority black cities have replaced long-serving black mayors with
white mayors.
And there are other signs of growing division in the black vote, perhaps
indications of growing frustration in the black community. In many cities,
what were once united black voting blocs have grown more divided. The
intra-group divide in cities such as Philadelphia and Memphis has often
centered on class lines (Adams 1994; Keiser 1997). In other cities, a range
of issues seems to have reduced black unity (Pohlman and Kirby 1996;
Wildstrom 1998). But in most cases, these divisions appear to have begun
with concerns over the ineffectual nature of black leadership.
Combined, these patterns in the black vote suggest that race, or at least
the race of candidates for political of¬ce, has lost some of its signi¬cance in
certain urban arenas. According to many political observers, rather than
focus on race, African Americans have begun to focus more on local eco-
nomic conditions and the quality of basic public services. David Bositis, a
researcher at the Joint Center for Political and Economic Studies, has con-
cluded that “Black voters have become more interested in getting results
than anything else” (Minzesheimer 1995). Newspaper reports focusing
on elections in largely black cities have also tended to conclude that race
is becoming less important: “Blacks in cities across America have reached
a frustration level and are looking for politicians that can change their
living conditions, regardless of color” (Ethnic News Watch 1995).
This sentiment has been echoed by at least some black voters. One black
voter in Gary expressed it this way: “My trash isn™t collected and there™s
been a giant pothole outside my door for two years. Do you think I care
about the color of the mayor™s skin?” (Tyson 1995). Another remarked:
“It™s not a white world. It™s not a black world. If you poke Scott King [the
white candidate], he™ll bleed the same color I do” (Minzesheimer 1995).
Changing White Attitudes

Finally, a Black Panther in Cleveland admitted, “We can™t vote based on
skin color, but based on what™s in our self-interest” (Gaither 1998).
Learning in both the white and black communities has arguably led to
elections in which race is peripheral “ or at least more peripheral than it
once was. An account of recent mayoral politics in Cleveland, for example,
concluded that “while race remains a factor . . . voters nevertheless appear
more willing to elect candidates of another race than at any time in the last
30 years” (Vickers 1997). A report from a 1995 election in Gary found that
“the city™s voters have mostly disregarded race in this campaign” (Tyson
1995). And in Baltimore, high crossover voting among both whites and
blacks has led many to maintain that race is losing its signi¬cance in urban
politics (Samuel 1999; Texeira 1999). Cleveland™s City Council president
described the trend as well as anyone else: “There are layers of issues in
any election, and while race is still an issue, it™s not the de¬nitive issue”
(Vickers 1997).
Certainly, there are many cities and elections where candidates™ race
continues to play a critical role and where African Americans continue
to expect great change from black elected of¬cials. But in other places
around the country, there appears to have been a real change in the black
community and in black political behavior. In these cities, blacks have
begun to realize that black leadership in and of itself cannot solve Amer-
ica™s racial problems. The letdown was probably inevitable. Many African
Americans had expectations that could never be ful¬lled by black incum-
bents. This is, in essence, white racial learning in reverse. Experience under
black leaders provided information that forced African Americans to rec-
ognize the limitations of black leadership. Although whites and African
Americans differ in many ways, they both appear to be open to at least
some change.

latino and asian american leaders
Another important case to consider in the context of American racial
politics is learning under Latino and Asian American incumbents. Demo-
graphic changes over recent decades have increasingly placed Latinos and
Asian Americans at the forefront of American politics: the number of
Latinos in of¬ce has more than doubled in the last two decades, so that
today there are over 4,500 Latino elected of¬cials nationwide, Latinos
occupy twenty-four seats in Congress, and one state has a Latino governor
(NALEO 2005). Asian American representation is also growing rapidly,
though from a much smaller population base. As one recent article put it,
Other Cases Where Information Could Matter 151

“Asian American representation in the hallways of power has gone from
barely noticeable to modestly in¬‚uential” (Ratnesar 1998).10 Between
1996 and 2000 alone, the number of Asian Americans holding of¬ce
increased by 10 percent nationwide (APALC 2001). In short, the growing
presence of these two groups is hard to ignore.
Should we expect the arrival of these new minority leaders to mark
the beginning of another learning process for white Americans? In many
ways, the context surrounding the arrival of Asian American and Latino
leaders mirrors the context surrounding the ¬rst African American elected
of¬cials. Most white Americans have had relatively limited contact with
members of either group, and few have witnessed either Asian Americans
or Latinos in positions of authority over the white community. There is
also evidence of fairly widespread and fairly negative stereotyping of both
groups. Whites are, for example, quite likely to believe that Latinos are
unintelligent, welfare prone, hard to get along with, and violent (Bobo
and Johnson 2000). Asian Americans, by contrast, tend to be viewed as
foreign and mysterious (Kim 1999; Lee 2000).11 Whites are also as likely
to admit feeling threatened by Latinos and Asian Americans as by African
Americans (Bobo and Johnson 2000). In particular, in places where either
group comprises a rapidly expanding or relatively large portion of the
population, whites have singled them out and acted as if they posed a
real threat (Hero 1998; Alvarez and Butter¬eld 2000; Hajnal, Gerber,
and Louch 2002). This suggests that many whites may have exaggerated
perceptions about what might happen under Latino or Asian American
Anecdotal evidence points in this same direction. In response to minor-
ity challenges, some white campaigns have warned of “an impending
[Latino] takeover” and, at least according to political observers, some
contests have “centered on the possibility of [Latino] political dominance”
(Rosales 2000: 94; Stuart 1983b). In other cases, the white community has
mobilized to try to prevent Latino or Asian American victory. In San Anto-
nio, for example, a Latino bid to gain control of the city council led to two
years of what has been described as “ethnic turmoil” (Stevens 1981). And
in Miami, when Cuban control of the County Commission became a pos-
sibility, other groups rallied in a series of highly charged elections to try to

10 Today there are over 600 Asian Americans at the federal, state, or local level, including
two senators and four members of the House of Representatives (APALC 2005).
11 Of all minority groups, it is Asian Americans with whom whites feel they have least in
common (Lee 2000).
Changing White Attitudes

prevent a Cuban takeover (Stuart 1983a). Another sign that whites have
real concerns about Latino and Asian American leadership is the fact that
Latino and Asian American campaigns often try to address those fears by
downplaying their race and ethnicity in a similar fashion to some African
American candidates. A range of successful Latino and Asian American
candidates have made claims similar to those of Jay C. Kim, who declared
to white voters: “I don™t have a special agenda for Asian Americans. They
should not expect anything special from me” (Mydans 1993). According
to many observers, Henry Cisneros “assiduously tried to build bridges”
in his effort to become the ¬rst Latino mayor of San Antonio (Stevens
1981). During his campaign for mayor in Denver, Cisneros often stated,
“the issue was not to elect a Hispanic candidate but to elect a candi-
date with a vision, with ideas for the city” (Schmidt 1983). Accounts of
Michael Woo™s bid for the mayoralty in Los Angeles also suggested that
he took pains not to identify too closely with Asian Americans or their
political interests (Clifford 1993). In order to win white votes, these new
minority candidates have, in the views of some, “become in effect ethnic
neutral” politicians (Mydans 1993).
And, as is the case with black leadership, the tenures of Latino and
Asian American leaders seem to have provided the white community with
positive or anti-stereotypical information. Since Latinos and Asian Amer-
icans have entered of¬ce only recently in most cases, we do not have
extensive information regarding the substantive impact of either Latino
or Asian American representation. But what we do have suggests that
Latino and Asian American elected of¬cials effect little change in the rel-
ative fortunes of whites and non-whites and at best deliver small gains to
some segments of the minority community. Fairly broad studies of Latino
representation in Congress and of Latino mayors and council members, in
particular, have revealed “few major departures from existing policy.”12

12 The quote is from Hero (1992: 152). Hero and Tolbert (1995) have found that the pol-
icy choices of Latinos in Congress were no different from those of whites of the same
party who represented similar districts. Similarly, Kerr and Mladenka (1994) could ¬nd
no real change in minority employment patterns under Latino mayors and council mem-
bers. Mladenka (1989) did, however, show some marginal changes in Latino employment
in places with greater Latino representation on the city council. More detailed analysis
of individual cases suggests that Latino leaders, just like their African American coun-
terparts, have often tried to enact policies that appease local businesses and middle-class
interests (Munoz and Henry 1997). Accounts of Henry Cisneros™s mayoralty are typ-
ical. Observers in San Antonio claimed that “Cisneros promoted neither his Mexican
American identity nor the speci¬c interests of that community. He acted as a mayor who
happened to be Mexican American” (Munoz 1994: 112). One of the toughest critics
referred to Latino representation in San Antonio as the “illusion of inclusion” (Rosales
Other Cases Where Information Could Matter 153

No systematic assessment of the policy effects of Asian American elected
of¬cials has been undertaken, but anecdotal evidence from a number of
Asian American of¬ceholders reveals little in the way of dramatic change.
Thus, there are reasons to expect that learning will occur under Latino
and Asian American incumbents.
At the same time, there are important differences between the black
context and the Asian American and Latino contexts that should alter the
information equation somewhat. Speci¬cally, at least two factors make
Latino and Asian American leadership less threatening than black lead-
ership. First, Asian Americans and Latinos generally run for of¬ce after
whites have experienced black leadership or at least have heard about
black incumbents elsewhere in the country. The learning that has occurred
under black leaders should help to reduce fears about Latino and Asian
American challengers. Second, the policy divide between white Ameri-
cans and the Latino and Asian American populations is not nearly as
sharp as the divide between white and black America (Kinder and Sander
1996; Uhlaner 2000; Hajnal and Baldassare 2001; Hochschild and Rogers
1999). The threat from Asian Americans could be particularly small for
at least three other reasons. The Asian American population is itself more
divided on most policy questions than either the Latino or African Amer-
ican communities (Tam 1995; Garcia 1997; Warren, Corbett, and Stack
1997; Hero 1998; Espiritu 1992). Asian Americans also have a socio-
economic status more similar to that of whites than have the other two
minority groups. And ¬nally, whites tend to hold some distinctively pos-
itive views of Asian Americans “ namely, that they are intelligent, hard-
working, and high achieving (Bobo and Johnson 2000; Lee 2000). Thus,
while I would expect the information model to apply to white Americans
experiencing Latino and Asian American leadership, I would also expect
learning and changes in white political behavior to be more muted.
Has experience with Latino and Asian American incumbents changed
white perceptions and white political behavior? Is there any evidence of
learning under Latino and Asian American incumbents? The answer to
both of these questions is a provisional yes. Although Latino and Asian
American of¬ceholding is very much a work in progress and the data
are limited, at ¬rst glance there does appear to be a pattern of changing
white behavior in response to experience with Latino elected of¬cials. The
evidence is clearer for whites who experience Latino leadership than it is
for whites who live under Asian American incumbents but in both cases
there are signs that white Americans are learning.
To mirror as best as possible the tests for change under African
American leadership, I assembled data on white voting patterns in Latino
Changing White Attitudes

challenger and Latino incumbent mayoral elections.13 Just as I did with
African Americans, I focused on the ¬rst Latino to win the mayoralty in
major American cities. As before, I compared white support for the Latino
challenger to white support for the same Latino running for the ¬rst time
as an incumbent. The number of cases here is small. There are only four
major cities that have elected Latinos who have then gone on to run
for reelection (Miami, San Antonio, Denver, and San Jose). The results,
however, are fairly clear. In three out of the four cases, when the same
candidate ran for reelection, white support increased appreciably. The
seven-point gain in white support (from 52 to 59 percent) that occurred
in San Jose was typical.14 Only Federico Pena in Denver faced as much
white opposition when he ran as an incumbent as he had when he ran
as a challenger. And even here, analysis of the white vote suggests that
the role of race diminished in Pena™s reelection bid; according to Hero
(1992), party and ideology played a more central role in that election.
Latino incumbents™ success has not been con¬ned to ¬rst reelection bids:
the four incumbents have won reelection in 13 out of 14 attempts; the loss
occurred in Miami, where Maurice Ferre lost not to a white challenger
but to another Latino, Xavier Suarez. We can also see evidence of declin-
ing turnout in these elections. San Antonio experienced one of the most

13 Estimates of the white vote come from newspaper accounts and published case studies
of these elections and are based on either exit polls or ward- or precinct-level analysis
(New York Times 1973; Munoz 1994; Hero and Beatty 1989; Gerston 1998; Scheibal
14 Interestingly, if the white vote in these cases is at all representative, it appears that white
voters are much more willing to support Latino mayoral challengers than they were
to support black mayoral challengers. The 52% average white support for Latino chal-
lengers stands in fairly sharp contrast to the comparable ¬gure for black challengers: 30%
white support. Asian American challengers, by almost all accounts, have been even more
successful at garnering white support. One study of California, for example, concluded
that Asian Americans “have been the most successful of minority candidates in winning
white votes” (Ratnesar 1998). The fact that the majority of Asian American elected of¬-
cials outside of Hawaii have won of¬ce in places where white voters predominate attests
to the ability of Asian American candidates to obtain fairly broad white support. In Cal-
ifornia, the state with the largest Asian American population, Cain and Kiewiet found
that 73% of Asian American elected of¬cials represented areas that were less than 10%
Asian American (1986). Gary Locke™s successful bid to become the mainland™s ¬rst Asian
American governor in a state that was only 5% Asian American, and Norman Mineta™s
victory in a congressional district that was only 2.5% Asian American are just two of the
more famous examples. The fact that party identi¬cation and political ideology mattered
as much or more than race/ethnicity in many of these challenger contests lends further
support to the notion that whites were less threatened by Latino and Asian American
candidates (Hero 1992). It may be that race and ethnicity are less of a hurdle for Latino
and Asian Americans than they are for African American challengers facing white voters.
Other Cases Where Information Could Matter 155

severe drops when turnout fell to 17 percent in 1983. But almost across
the board, interest in mayoral elections has declined as residents gained
more experience with Latino mayors (New York Times 1983b; Kaufman
It is dif¬cult to tell whether these trends are spreading to other cities.
More Latino candidates seem to be winning of¬ce in minority Latino
places. Recent victories by Gus Garcia in Austin, Eddie Perez in Hartford,
and Judith Valles in San Bernardino “ all minority Latino cities “ are
suggestive of a trend. Antonio Villaraigosa™s victory in racially diverse
Los Angeles could represent another important turning point.15 There
is also some indication that over time whites are increasingly willing to
support Hispanic candidates. Studies of certain states have found that “the
number of crossovers is increasing” (Suro 1991). Although some highly
racialized elections continue to occur, the overall pattern may point to
a lessening of white racial concerns and declining white opposition to
Latino candidates.16
The data on Asian American incumbents are more limited. In large
part, this is because of the small number of cases of Asian American
leadership in major of¬ces. Asian Americans have, for example, won the
mayoralty of only two cities with populations of over 200,000, and esti-
mates suggest that not much more than a handful of Asian Americans
have been directly elected to mayoral of¬ces around the country (APALC
2003). Nevertheless, it is clear that Asian American incumbents who win
of¬ce in primarily white areas tend to do extremely well when they run
for reelection. Analyzing available data on Asian American candidates, I
found that across the country, 98 percent of Asian American incumbents
who ran for reelection at the city council level or higher between 2000 and
2003 won their reelection bids.17 And at least at the congressional level,
growing victory margins imply that over time Asian American incum-
bents are winning with an increasing greater amount of white support.18

15 In 2005, there were 21 Latinos who were serving as mayors of cities with a population
over 100,000 (NALEO 2005).
16 One of the most prominent recent cases of polarized voting was in San Antonio, where
Maria Berriozabal lost a mayoral bid with 95% of the Mexican vote and only 20% of
the white vote (Munoz 1994).
17 These results are derived from election results reported by the Asian Paci¬c American
Institute for Congressional Studies. The data set is far from complete (especially for of¬ces
below the level of the state legislature), but it does show that of 60 Asian American
incumbents who sought reelection over this time period, 59 were successful.
18 With few exceptions, congressional election returns since the 1990s indicate that Asian
American candidates have generally garnered more and more support as their time in
Changing White Attitudes

Moreover, few of these reelection bids appear to stimulate much in the
way of a debate over race. Gary Locke, for example, “cruised to a second
term” as governor of Washington state in a campaign that featured little
attention to race. In a pattern similar to that of black incumbents, Locke
simply ran on his record (Pierce 2000). Finally, although Asian Amer-
icans have almost always had to win over non-Asian American voters
in order to get elected, the number of Asian Americans who are able to
do so successfully is clearly growing over time. Much of the growth in
Asian American representation can be directly attributed to the growing
Asian American population, but white voters also appear to be playing
an increasingly important role.
None of these trends provides direct evidence of white learning under
Asian American or Latino incumbents. But the available evidence is both
interesting and suggestive. The patterns that we see in the vote and in the
nature of Latino and Asian American incumbent campaigns hint at grow-
ing white acceptance of Latino and Asian American leadership. Clearly,
a plausible reading of the data is that white Americans are learning from
their experiences under Latino and Asian American incumbents. If true,
this suggests that the information model can be applied across a fairly wide
range of cases. It also suggests that Latino and Asian American leadership
may be another important learning tool that could lead to positive change
in inter-group dynamics.

transitions around the world
The patterns that we see across all of these different cases raise questions
about the applicability of the information model to a wider range of
political contexts around the world. The transition to democracy and
majority black rule in South Africa, for example, seems to ¬t many of
the criteria of the model. White fears were widespread in South Africa
before the election of the African National Congress, and the transition
to black control did not lead to a wholesale redistribution of resources
or the downfall of the white community. Thus, the possibility of white
learning in South Africa certainly exists.19

of¬ce increased. Since most of the Asian American incumbents are running in primarily
white districts, this rules out a white backlash explanation and implies that whites are
becoming more supportive over time.
19 Although there is little evidence of a dramatic shift in white attitudes in response to black
leadership in South Africa, the results of the most recent national elections in 2004 did
seem to demonstrate a slight reduction in racial divisions (Ferree 2005).
Other Cases Where Information Could Matter 157

There may be other contexts beyond the arena of race and ethnicity,
moreover, where the information model is applicable. There is nothing in
the information model itself that requires a racial divide or racial stereo-
types. All that is required is a lack of information about one particular
group and fear about what that group will do once in power. Thus, the
information model can be seen as a model for understanding transitions of
power from once-dominant groups to underrepresented minority groups.
One could apply such a model to the cases of religion or gender. Did the
election of John F. Kennedy as the ¬rst Catholic president lead to a change
in attitudes toward Catholic candidates? Certainly, one could make a rea-
sonable claim that America™s experience under Kennedy reduced the sig-
ni¬cance voters accorded to the Catholicism of candidates. The concerns
about a Catholic president that were expressed when Kennedy ran for
of¬ce seem not to have played an obvious role in subsequent elections
involving Catholic candidates around the country.
Whether gender transitions provide the same information and have
the same effect on attitudes and voting behavior is less clear. Interactions
between men and women are generally much more extensive than inter-
actions between different religious or racial groups. As a result, lack of
information may not be as great a problem when female candidates run
for of¬ce for the ¬rst time. Although existing studies do suggest that men
hold stereotypes about female candidates, researchers have found that
these stereotypes are not universally negative. Moreover, it is not clear
that stereotypes and expectations surrounding female candidates always
reduce the level of support that female candidates receive (Huddy and
Terkildsen 1993; McDermott 1997). My suspicion is that learning, to the
extent it occurs under women in of¬ce, is so muted that it would be dif¬-
cult to register using the blunt measures available to researchers. On the
other hand, it is not inconceivable that there was at least a slight adjust-
ment to gender stereotypes in the wake of Margaret Thatcher™s strong-
willed leadership as the ¬rst female prime minister of England. Thus, in
exceptional cases, electing women to of¬ce might spark real change.
In many of the cases that I have highlighted, and in a range of other
political transitions around the world, the main barrier to one group
ceding power to another group is thought to be historical enmities and
irreconcilable differences over race, religion, or ideology. But these same
con¬‚icts can just as easily be viewed as being driven by uncertainty and a
lack of information. This may give us a very different perspective on polit-
ical transitions. For groups losing power, it may not be ideological divides,
racial hatreds, or religious antagonisms that really drive their actions and
Changing White Attitudes

ultimately lead to mobilization and violence. Instead, from the perspec-
tive of the information model, the root of the problem may be a lack of
information. Perhaps it is impossible to provide information to reassure
formerly dominant groups. Indeed, transitions, when allowed, sometimes
have the very consequences that members of formerly dominant groups
fear. But when both groups have essentially the same interests “ the main-
tenance of the economic health of the larger community and the continued
vitality of the country or region “ there may be room for information to
play a role. In these cases, if the right information can be provided, it
may help to overcome fears driven by religious, ideological, or racial dif-
ferences. At the very least, it seems worthwhile to ask deeper questions
about the role of information and of uncertainty in con¬‚icts around the
A Tale of Caution and Hope

For years, African Americans fought for access to the vote and the ability
to choose their own leaders. When that opportunity ¬nally arrived, many
in the African American community were justi¬ably jubilant. Over the
ensuing years, black elected of¬cials have tried to live up to the hopes
and expectations of the black community. In many cases, they have
tried valiantly to improve conditions in the black community, and fre-
quently their efforts have been rewarded. According to a range of studies,
black representation has made a difference. Nevertheless, to many if not
most observers of American politics, the victory seems somewhat hol-
low, as black elected of¬cials simply have not been able to alleviate basic
problems facing the African American community. Poverty, poor edu-
cational achievement, crime, racial segregation, and other fundamental
racial inequalities remain largely unchanged despite years and in some
cases decades of black leadership. In light of the ongoing problems plagu-
ing the African American community, some critics now claim that black
leadership is “largely irrelevant.”1
One of the goals of this book has been to show that this view represents
only part of the story of black representation. Existing studies have failed
to uncover one of the most important consequences of black representa-
tion because they have largely focused on the effects of black leadership on
the black community. By examining the effects of black leadership on the

1 The quote is from a book entitled We Have No Leaders in which Robert Smith lambasts
the conservative policies of African American leadership. Other critics, such as Adolph
Reed, have argued that black leadership often “doesn™t really amount to a transition to
local black rule” (1989: 44).

Changing White Attitudes

white community, I have shown that black of¬ceholding has an important
impact on the attitudes and actions of white Americans. The tests that I
have discussed here reveal change on a number of fronts. Under black
mayors, there is a measurable shift in white voting patterns and in the
racial sentiments expressed by a portion of the white population. Black
representation appears to diminish the role of race in the vote, to boost
white support for black candidates, relax racial tension, and foster more
positive white attitudes toward both black leadership and the black com-


. 5
( 7)