Race and Gender: An Analysis of the Sources and Reporters in the Networks'
Coverage of the Year 2000 Presidential Campaign
Geri Alumit Zeldes and Frederick Fico
Michigan State University
Geri Alumit Zeldes, Ph.D.
Visiting Assistant Professor
School of Journalism, Michigan State University
348 Communication Arts & Sciences
East Lansing, MI 48824-1212
E-mail: [log in to unmask]
Fred Fico, Ph.D.
School of Journalism, Michigan State University
386 Communication Arts & Sciences
East Lansing, MI 48824-1212
E-mail: [log in to unmask]
Submitted to the Radio-Television Journalism Division for presentation at
the 2003 AEJMC Convention.
ABSTRACT: Race and Gender: An Analysis of the Sources and Reporters in
the Networks' Coverage of the Year 2000 Presidential Campaign
More than 50 years after Hutchins' Commission on Freedom of the Press and
more than 30 years after the Kerner Commission, American discourse still
focuses on racial disparities and the lack of minority representation in
Decades later, our content analysis of 333 campaign stories broadcasted by
ABC, CBS and NBC during the 2000 presidential election found minimal
representation of minorities and women as reporters and sources. Stories
by women and minority reporters were more likely to use and give time to
women and minority sources than did stories by male and Caucasian reporters.
The Race and Gender of Sources and Reporters in the Networks' Coverage of
the Year 2000 Presidential Election
More than half of likely voters regularly obtain information about
political campaigns from either network or local television news coverage
("News Source of Choice," 2000). Given the high visibility of national
elections, reporters see election reporting as one of their central
missions. To what extent do they fulfill social responsibility norms,
which emphasize that reporters as gatekeepers provide citizens with
information from "all sides" to guide election voting? In other words, do
reporters seek out sources that represent a diversity of race, gender and
opinion to allow voters to make an educated decision?
This study examined race and gender representation of reporters and sources
on ABC, CBS and NBC during the year 2000 presidential election, and related
coverage to race, gender and other source-type diversity. The goal of such
an analysis is to assess whether arguments to diversify set forth by the
Hutchins' Commission and Kerner Commission several decades ago are still valid.
Are we still "separate and unequal" in the news?
In 1942, Time magazine founder Henry Luce asked former Yale classmate
Robert Hutchins to head a blue-ribbon panel called the Commission on
Freedom of the Press to explore mounting problems facing the press such as
the concentration of mass media ownership. In 1947 the group released a
report that set forth a social responsibility code that required the press
to provide five basic services. The third requirement obligated the press
to show a "representative picture of the constituent groups in
society." Peterson (1956) further explicated,
this requirement would have the press accurately portray the social groups,
the Chinese and the Negroes, for example, since persons tend to make
decisions in terms of favorable or unfavorable images and a false picture
can subvert accurate judgment (p. 91).
Two decades later, riots erupted in Los Angeles, Newark, New Jersey and
Detroit during the first half of the 1960s. In response, President Lyndon
B. Johnson appointed Illinois Gov. Otto Kerner to chair a commission of
business, political and civil rights leaders to investigate the nation's
ethnic tensions. In 1968 the commission issued a "Report of the National
Advisory Commission on Civil Disorders" that concluded that racism and
economic inequality sparked the riots and that if racial divisiveness
continued, the country would have two societies, "one black, one white –
separate and unequal" ("30 years after," 1998).
Specific to the news content, the Kerner Commission chided news media
executives with "flaunting an affluent white society before nonwhite
Americans and called for immediate reform in what was termed 'shockingly
backward' hiring practices" (Ziegler and White, 1990, p. 216). The report
warned that an absence of reform would encourage minorities to associate
mass media as an instrument of white elites.
Nearly 10 years later in "Window Dressing on the Set," the U.S. Commission
on Civil Rights (1977) noted that responses to the 1968 report resulted in
hiring more minorities in the newsroom and implementing affirmative action
programs, but as of the mid-1970s, the numbers of minority reporters at the
network level were still low. Stone (1988) found that the numbers of
minorities in TV newsrooms increased through the 1970s and 1980s and hit a
plateau in the mid-1980s.
But by 1991, the Radio Television News Directors Association reported that
the numbers of women and minorities on TV news staffs had remained at a
standstill, with women at a third and minorities at 17.4 percent of TV news
staffs ("Media Report," 1992). In 2001, Papper's survey for the RTNDA
found that women made up 38.6 percent and minorities 20.6 percent of the TV
news workforce, down from 24.6 percent in the previous year, but higher
than a decade earlier ("Women and " 2002). The numbers of news staff are
coming closer to representing the population: Census 2000 figures show
that women make up 50.9 percent of the population whereas minorities made
up 24.9 percent of the population (U.S. Census Bureau, 2001a; U.S. Census
But does this increased presence of minorities and women on the network
level result in a more representative picture of society sought by the
Hutchins Commission and other concerned groups? Do the networks, more than
30 years after the Kerner Commission, still present news through a
predominantly white and male filter? Does having women and minority
reporters change this filter and facilitate a broader reflection of such
groups in news reports? The present research can provide some answers.
Content analysis of news on the network level has found minimal
representation of women and minorities as reporters and sources on network
news, despite the growing minority population (Cann and Mohr, 2001; Liebler
& Smith, 1997; McClellan, 1998;Whitney, Fritzler, Jones, Mazarella & Rakow,
1989; Ziegler and White, 1990). Ziegler and White examined a sample of the
network newscasts broadcasted in 1987 and 1989 and found that white
reporters covered 93 percent of the 332 stories and 88 percent of the
reporters were men. In both samples, NBC had the lowest degree of
representation with 97 stories reported by white men, 8 by white women and
no non-white reporters.
In Andrew Tyndall's study for the Women, Men and Media Project that
analyzed the first six months of network newscasts in 1998, he found that
CBS had the lowest degree of female representation. ABC had the highest
representation with 32 percent of all reports filed by women in 1998, up
from 25 percent in 1996. On NBC, women filed only 24 percent of the
reports, up from 17 in 1996 and on CBS, only 11 percent of the reports were
filed by women, down from 15 percent in 1996. Overall, women filed 22
percent of the reports, up from 19 percent in 1996 and 14 percent in 1988
A survey conducted by the Center for Media and Public Affairs in 1998 found
that overall, 16 of the 20 most visible reporters on network evening news
were white males ("Women, Minority," 1999). The center's study however,
found results similar to Tyndall's study that show the proportion of
stories reported by women and minorities increased in 1998 from
1996: Overall, women reported 19 percent and minorities, African-American,
Hispanic, Asian and other ethnic backgrounds, reported 10 percent of the
stories. For three consecutive years, ABC led the networks with the
highest proportion of women on ABC with 25 percent, up from 23 percent in
1996 and the highest proportion of minority representation with 12
percent. NBC had the greatest increase of women with 17 percent, up from
10 percent in 1996 and the second highest minority representation with 10
percent, and the proportion of women represented on CBS fell to 14 percent,
down from 15 percent in 1997, but minority representation increased from 1
percent in 1997 to 7 percent in 1998.
Ziegler and White (1990) also found that of the 1461 sources used, 72
percent of the sources were white men, 9 percent were white women, 15
percent were nonwhite men and 2 percent were nonwhite women; by gender, 88
were men and 12 percent were women. Tyndall's study found that 87 percent
of the expert sound bites were from men and 92 percent were from white
sources. Only 13 percent of the sound bites were from women and 6 percent
were from minorities (McClellan, 1998). Cann and Mohr (2001) found men
were over-represented as presenters, reporters and expert sources on five
Australian networks and like Tyndall, Liebler and Smith (1997), they found
that male sources were more likely to be shown as experts and women as
non-experts or ordinary people.
In "The Black Image in the White Mind," Entman and Rojecki (2000) show that
African-Americans are rarely presented as experts in the news and are more
likely interviewed as sports and entertainment figures, victims of
discrimination, and criminals. The one exception is on black issues, where
African-Americans are more likely than whites to be used as
experts. Finally, Liebler and Smith found that reporter gender did not
impact news source gender; reporters of both sexes tended to use more male
Research Questions and Hypotheses
Previous research has documented, separately, trends in demographic and
other characteristics of both journalists and their sources. But the
underlying assumption of much of the concern for diversity is that
diversified newsrooms are necessary for news stories that appropriately
present a diverse society and its perspectives. We first pose the
following research questions that address race and gender reporter
qualities and gender and race source qualities:
RQ1: What was the percentage of network news stories by female reporters in
the year 2000 presidential campaign?
RQ2: What was the percentage of network news stories by minority reporters
in the year 2000 presidential campaign?
RQ3: What was the percentage of network news stories in which women were
RQ4: What was the percentage of network news stories in which minority
group members were sources?
Moreover, if the premise of newsroom diversity is correct, then the
following hypotheses should be supported:
H1: Female reporters will use women sources in a greater percentage of
their stories then will male reporters.
H2: Female reporters will give women sources more time in their stories
than will male reporters.
H3: Female reporters will use sources such as experts and ordinary people
in a greater percentage of their stories than will male reporters.
H4: Minority reporters will use minorities in a greater percentage of their
stories then will Caucasian reporters.
H5: Minority reporters will give minority sources more time in their
stories than will Caucasian reporters.
H6: Minority reporters will use sources such as experts and ordinary people
in a greater percentage of their stories than will Caucasian reporters.
This study explores these questions with a content analysis of the universe
of stories broadcasted on the evening newscasts of ABC, CBS and NBC during
the year 2000 presidential campaign. These broadcasts were taped from the
Labor Day start of the campaign to November 6, the day before Election
Day. Election segments were identified based on explicit graphic symbols
used by the networks, always accompanied by verbal transitions by
anchors. Such segments could consist of one or more stories. Individual
stories were often read by an anchor, and were frequently accompanied by
visual material and sound bites from the campaign. Most commonly, however,
anchors introduced stories from reporters who covered the Bush and Gore
campaigns. Race and gender of reporters were recorded for each story.
Each story in a multi-story segment could therefore be easily identified by
the transition to a separate journalist. The content analysis focused
explicitly on the use of personal sources in these stories. Such sources
of course, included George Bush, Al Gore and their running mates. But up
to six non-candidate sources were also coded for each story.
The researchers coded the length of each source's sound bite, his or her
gender, race and source type. Length of sound bite was recorded in seconds
and gender was coded as male or female. Race was coded as Asian, African
American, Caucasian, American Indian and Alaska Native, or Native Pacific
Source types were defined as the following:
1. Partisan: Sources who made explicit assertions supporting Bush, Gore,
Nader and Buchanan.
2. Horse Race Expert: Sources identified from academe, polling or
consulting companies, who did NOT make explicit candidate endorsements and
who were explicitly cited for their credentials to assess candidate
prospects for success.
3. Issue Expert: Sources identified from academe, government or interest
groups, who did NOT make explicit candidate endorsements and who were
explicitly cited for their credentials to provide interpretation or context
for election issues.
4. Ordinary Person: Sources who did NOT make explicit candidate
endorsements nor have horse race or issue expert status.
Reliability of Measures
The coding procedure employed in this study initially had one researcher
and two research assistants identify election-relevant stories and another
researcher validating that judgment. Two researchers coded the above
source and reporter characteristics in the stories. A coder reliability
assessment was made on about 10 percent of the relevant randomly sampled
election stories. Percentage of agreement on the variables relevant for
this analysis ranged between 81 percent and 100 percent. Scott's Pi
computations to correct for chance agreement ranged from .78 to 1.0.
Questions and hypotheses posed in this study are addressed with story
proportions in which non-candidate sources were used and with means for
their sound bite lengths. All stories on the 2000 election from designated
evening newscasts were assessed, so inferential statistics are not
necessary for generalization.
The researchers coded 333 election-relevant stories, which were almost
evenly distributed among the three networks: ABC aired 115, CBS aired 105
and NBC aired 113.
CBS had almost all (97%) of its stories reported by male reporters,
followed by ABC (88 %) and NBC (53%). ABC and NBC had almost all, 99
percent of their stories, covered by Caucasian reporters. CBS had the
highest percentage of its stories (28%) covered by minority
reporters. Female sources were used in 23 percent of CBS stories, in 24
percent of ABC stories and in 31 percent of NBC stories. Minority sources
were used in 7 percent of CBS stories, in 9 percent of ABC stories and in 7
percent of NBC stories.
Research Question Results
RQ1 asked about the percentage of stories by female reporters on network
news stories of the year 2000 presidential campaign. Female reporters
covered 70 (21%) of the 333 stories and male reporters covered 263 (79%) of
the stories (See Table 1). Therefore, male reporters covered nearly
four-times more stories than female reporters.
RQ2 asked about the percentage of stories by minority reporters on network
news stories of the year 2000 presidential campaign. African American
reporters covered 31 (9%) of the total stories and Caucasian reporters
covered the rest, 302 (91%) stories (See Table 2). The researchers did not
code any of the reporters as Asian, American Indian and Alaska Native, or
Native Pacific Islander.
RQ3 asked about the percentage of stories in which women were
sources. Women were sources in 159 (52%) of the stories.
RQ4 asked about the percentage of stories in which minority group members
were sources. Minorities were sources in 25 (8%) of the stories.
H1 stating that female reporters will use women sources in a greater
percentage of their stories then will male reporters was supported. Female
reporters used both male and female non-candidate sources in a higher
percentage of their stories than did male reporters (See Table 1). And
female reporters used women sources in nearly twice the percentage of their
stories than did male reporters in their stories.
H2 stating that female reporters will give women sources more time in their
stories than will male reporters was supported. Female reporters gave
female sources about two seconds more time in stories than did male
reporters (See Table 1). On average, however, female reporters gave more
time to male sources than they did to female sources.
H3 that stated female reporters will use expert sources in a greater
percentage of their stories than will male reporters was supported. As
presented in Table 3, female reporters used more partisan, horse race,
issue and ordinary persons in a higher percentage of their stories than
H4 stated that minority reporters will use minorities in a greater
percentage of their stories then will Caucasian reporters was
supported. Minority reporters used minority sources twice as much as
Caucasian reporters did (See Table 2). Minority reporters also used
Caucasian sources in a greater proportion of their stories than did
H5 stated that minority reporters will give minority sources more time in
their stories than will Caucasian reporters was supported. Minority
reporters gave minority sources more time although this time averaged less
than two seconds (See Table 2). Moreover, African American reporters when
compared to Caucasian reporters gave minority sources time in a higher
percentage of their stories and longer sound bites; the time given to
minority sources however, appears especially negligible when considering
how much time reporters in general gave to Caucasian sources. African
American reporters actually gave Caucasian sources more than five seconds
longer to talk then Caucasian reporters gave Caucasian sources.
H6 stated that minority reporters will use sources such as experts and
ordinary people in a greater percentage of their stories than will
Caucasian reporters was partially supported. As presented in Table 4,
minority reporters used only horse race and issue expert sources in a
higher percentage of their stories than Caucasian reporters.
Conclusions and Implications
Our content analysis of network news coverage of the year 2000 presidential
campaign might disappoint members of the Kerner Commission and others who
decried the media's tendency to "flaunt an affluent white society before
nonwhite Americans." In our study, Caucasians and men were
over-represented as reporters and sources on network news, replicating
earlier research on network news coverage. More than a decade ago, Ziegler
and White (1990) in their study of network newscasts in 1987 and 1989,
found that Caucasian reporters covered 93 percent of the 332 stories and 88
percent of the reporters were men. Tyndall's study showed that Caucasian
reporters covered 90 percent of the stories and 78 percent of the reporters
were men (McClellan, 1998). A survey conducted for the Center for Media
and Public Affairs showed that in 1998, Caucasian reporters covered 90
percent of the stories and 81 percent of the reporters were men ("Media
Report," 1999). In this study we found that Caucasian reporters covered 91
percent of the 333 stories and 79 percent of the reporters were
men. However, while the survey for the center found African-American,
Asian, American Indian and Alaska Native or Native Pacific Islander
reporters, we only found African-American reporters. Our results are even
more disconcerting when considering the growing numbers of
minorities: Census 2000 figures show that one out of two people in the
United States are women and one out of four identify themselves as
belonging to a minority group (U.S. Census Bureau, 2001a; U.S. Census
Our study also confirmed Tyndall and the center's findings regarding the
under-representation of female reporters across networks. What is
disturbing is that our findings skewed even further toward men. Like
Tyndall and the center's studies, we found that CBS had the lowest degree
of female representation among the networks. Tyndall found that only 11
percent of CBS' stories and the center found that women filed only 14
percent of CBS's stories; we found that women filed only 3 percent of CBS'
stories. On ABC, Tyndall found women filed 32 percent of all reports and
the center's study found women covered 25 percent of the stories in 1998;
we found only 12 percent of the stories filed were from female
reporters. However, we differed dramatically on our findings on
NBC. Tyndall found that women filed 24 percent of the reports and the
center's survey found women filed 17 percent of the reports, and we found
that women filed 47 percent of the stories.
This study also replicated Liebler and Smith's (1997) results in that
reporters of both sexes tended to use male sources in more stories and gave
them more time to speak per story than female sources. Female reporters
also gave male sources on average four more seconds than female sources and
gave female sources two more seconds per sound bite. Liebler and Smith
gave two explanations for their findings:
One is that organizational and extra-organizational levels are still more
influential than the individual in deciding who and what constitutes
news. The other related explanation is that women have been socialized
into "male" definitions of newsworthiness and report accordingly.
These findings challenge the assumption that the hiring of more women and
minorities result in a more diverse reflection of society. Some findings
also suggest that female reporters on the network level have more of a bias
toward male sources and Caucasian sources than male reporters. More
research, both quantitative and qualitative, needs to be conducted to
discover if and why this is so before suggestions are made to discard
increased hiring of minorities as a solution to the lack of diversity in
Our study did however, show that women and minorities reported stories
using a broader mix of non-candidate sources of all types than did stories
by male and Caucasian reporters. Indeed when it came to reporting the 2000
election, male and Caucasian reporters were more likely to focus
exclusively on the candidates themselves, even to the exclusion of other
types of sources. This content study cannot illuminate why this should be
so. Possibly women and minority reporters were explicitly assigned to do
"broader" campaign stories in which different sourcing would be
necessary. Focused interviews with reporters and even participant
observation of their news coverage activities may help explain the patterns
found in this study.
The results of this study also have practical relevance in the continuing
discussion about newsroom and news story diversity. Since the Hutchins
Commission, discussion on diversity issues has led to efforts to hire
females and minorities in part to create a more diversified and news
product. More creative methods to increase diverse voices and faces as
sources in the news emerged in recent years. For example, ABC news
managers amassed a database of 480 minority sources, and news staff
managers have told employees that their job performance evaluation will in
part rely on the number of sources in this database that they contact. CBS
and NBC report similar requirements ("ABC ties," 2002). While this
research cannot address the social value merits of increased hiring of
female and minority reporters, it does call for more examination of how
hiring decisions affect news coverage. Debates on race and gender in the
news media must themselves take place in a broader context that includes
newsroom norms and values. Research can assess how well diversity is
achieved. But journalists must first define what they mean by that word.
Table 1. Gender of reporter and gender of non-candidate source by %, time
and Avg. sound bite length
Time given to non- Gender of Reporter
candidate sources Male (n=263) Female (n=70)
Sound bites in % of stories
Male 43% 67%
Female 22% 41%
Total seconds 2573 1175
Male 67% 65%
Female 33% 35%
Avg. sound bite length
Male 6.55 10.90
Female 3.23 5.87
M/F Diff. 3.32 5.03
Table 2. Race of reporter and race of non-candidate source by %, time and
Avg. sound bite length
Time given to non- Race of Reporter
candidate sources Afr. Am. (n=31) Caucasian (n=302)
Sound bites in % of stories
Minority 16% 6%
Caucasian 80% 51%
Total seconds 510 3232
Minority 8% 6%
Caucasian 92% 94%
Avg. sound bite length
Minority 1.30 .63
Caucasian 15.20 10.10
Min./Cauc. Diff. -13.9 -9.47
Table 3. Gender of reporter and frequency of type of non-candidate source
Type of non- candidate Gender of Reporter
sources Male (n=263) Female (n=70)
Partisan Source 30% 43%
Horse Race Expert 9% 33%
Issue Expert 11% 20%
Ordinary Person 17% 27%
Table 4. Race of reporter and frequency of type of non-candidate source
Type of non- Race of Reporter
candidate sources African Am. (n=31) Caucasian(n=302)
Partisan 61% 30%
Horse Race Expert 19% 13%
Issue Expert 16% 13%
Ordinary Person 19% 19%
ABC ties evaluations to use of diverse sources. (2002, May). The Quill,
90, 4: 33.
Cann, D. J., and Mohr, P. B. (2001). Journalist and source gender in
Australian television news. Journal of Broadcasting & Electronic Media,
Liebler, C. and Smith, S. J. (1997). Tracking gender differences: A
comparative analysis of network correspondents and their sources. Journal
of Broadcasting & Electronic Media, 41, 58-68.
McClellan, S. (1998, October 26). The whitemaling of network
news. Broadcasting & Cable, 128, 30-31.
News source of choice. (2000, February 14). USA Today, p. 1A.
Peterson, T. (1956) The social responsibility theory of the press. In Four
Theories of the Press. Urbana, IL: University of Illinois Press.
RTNDA research document little change for women, minorities. (Fall
1992). Media Report to Women, p. 4.
Report of the national advisory commission on civil
disorders. (1968). New York Times Edition. New York: Dutton.
Stone, V. A. (1988). Trends in the status of minorities and women in
broadcast news. Journalism Quarterly, 65 288-293.
30 years after Kerner report, some say racial divide wider. (1998, March
1). CNN Interactive, [On-line]
U.S. Census Bureau. (2001a, September 10). Male population grew faster
than female population Census 2000 analysis
U.S. Census Bureau. (2001b, August). The White Population: 2000: Census
U.S. Commission on Civil Rights. (1977). Window dressing on the
set: Women and minorities in television. Washington, DC: Author.
Whitney, D.C., Fritzler, M., Jones, S., Mazzarella, S., & Rakow, L.
(1988). Geographic and source bias in network television news
1982-1984. Journal of Broadcasting and Electronic Media, 33, 159-174.
Women and minorities in news are decreasing. (2000, December). USA Today,
Women, minority network reporters still struggling with visibility. (1999,
Spring). Media Report to Women, 27, 2: 1-3.
Zeigler, D. & White, A. (1990). Women and minorities on network
television news: An examination of correspondents and newsmakers. Journal
of Broadcasting & Electronic Media, 34 (2), 215-223.