05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
RACE TO JUDGMENT: STEREOTYPING
MEDIA AND CRIMINAL DEFENDANTS*
ROBERT M. ENTMAN**
K
IMBERLY A. GROSS***
I
I
NTRODUCTION
The media’s coverage of the Duke lacrosse story generated controversy
from the very beginning. Early on, criticism came from those who felt that the
media mistreated the accuser; later, critics wondered why coverage failed to
direct more attention to the weakness of the prosecution’s case. And
throughout, critics suggested that the media mistreated the accused, rushing to
judge them, abandoning the credo of innocent until proven guilty. What
happened to these young men was indubitably a travesty of justice, and there
are important lessons to be drawn regarding how standard media routines and
practices can undermine the presumption of innocence. However, these lessons
do not support critiques that trace media derelictions to “liberal bias” or
“political correctness.”
1
Nor should coverage as a whole be characterized as
consistently slanted against the defendants. Just as it was unfair to jump to the
conclusion that the prosecutor’s emphatic insistence on the defendants’ guilt
Copyright © 2008 by Robert M. Entman and Kimberly A. Gross.
This Article is also available at http://law.duke.edu/journals/lcp.
* An earlier version of this article was presented at the symposium conference, The Court of
Public Opinion: The Practice and Ethics of Trying Cases in the Media, at Duke University School of
Law, September 28–29, 2007. The authors would like to thank Gerard Matthews and Eric Walker for
research assistance.
** J.B. & M.C. Shapiro Professor of Media and Public and International Affairs, School of Media
and Public Affairs, The George Washington University.
*** Associate Professor of Media and Public Affairs, School of Media and Public Affairs, The
George Washington University.
1. For examples of such critiques, see
generally NADER BAYDOUN & R. STEPHANIE GOOD, A
RUSH TO INJUSTICE: HOW POWER, PREJUDICE, RACISM, AND POLITICAL CORRECTNESS
OVERSHADOWED TRUTH AND JUSTICE IN THE DUKE LACROSSE RAPE CASE (2007); STUART
TAYLOR, JR. & KC JOHNSON, UNTIL PROVEN INNOCENT: POLITICAL CORRECTNESS AND THE
SHAMEFUL INJUSTICES OF THE DUKE LACROSSE RAPE CASE (2007). See also Durham-in-
Wonderland, The Times: No Harm, No Foul, http://durhamwonderland.blogspot.com/2007/04/times-
no-harm-no-foul.html (Apr. 23, 2007, 00:01 EDT), in which KC Johnson makes the following statement
that typifies many critics’ analyses: “[T]he Times’ failure in the lacrosse case . . . was attributable to
reporters and editors allowing their worldviews to distort the facts.” As far as we can tell there is no
published book or article that engages in systematic content analysis of media coverage over the entire
period of the case. A number of sources provide qualitative analysis of selected media coverage. See,
e.g., T
AYLOR & JOHNSON, supra (focusing mainly on coverage that suggests a bias against the
defendants); David J. Leonard, Innocent Until Proven Innocent, 31 J. S
PORT & SOC. ISSUES 25 (2007)
(reaching a different conclusion by focusing on efforts to exonerate and defend the students).
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
94 LAW AND CONTEMPORARY PROBLEMS [Vol. 71:93
meant they were guilty, so it is wrong to assess the media by relying on
unsystematic impressions and inevitably flawed memories filtered through
stereotypes of, and generalized dissatisfaction with, U.S. journalism.
To draw the proper lessons from this incident, its coverage is best examined
in the larger context of research on media, race, and crime. Thus, this article
begins by reviewing the extensive literature in the subfield, highlighting the
ways crime news typically reinforces stereotypes of blacks and other nonwhites,
particularly Latinos, and contributes to more-generalized racial antagonism.
Understood in this context, news of the Duke lacrosse case diverges markedly
from typical crime coverage. This is worth bearing in mind when recalling the
amount of attention devoted to the injustices suffered by the Duke students
2
relative to the attention given to potential injustices suffered by more-typical
defendants.
This article systematically analyzes coverage of the Duke lacrosse case over
the year that it received attention in the news. It distinguishes between news
reporting and news commentary and between national and local coverage. In
analyzing the reporting, it also distinguishes between two phases. The first
phase, between March 14 and April 10, 2006, was before the first exonerating
DNA evidence became available. The post-DNA period, from April 10, 2006,
through April 11, 2007, the date on which the prosecutor announced that all
charges were being dropped, constitutes the second phase. The initial DNA
information could have clearly marked the unraveling of the case, serving as a
signal to reporters to dig deeper, but the weakness of the prosecutor’s case was
not fully reflected in much of media treatment until well into the post-DNA
phase. That is not the whole story, however. The data reveal a more
complicated pattern of news coverage that neither continually violated, nor
perfectly conformed to, journalistic ideals. News of this case is explained using
theories of news slant that emphasize how the skill and incentives of players in a
two-sided framing contest—here, criminal defense versus law-enforcement
agencies—interact with the incentives, norms, and limitations of the news
media. These theories take the analysis far beyond any simplistic, monocausal
explanation of media shortcomings in this single instance and toward an
understanding of the more general problems posed by normal media behavior
for the impartial administration of justice in the United States.
II
J
OURNALISTIC ROUTINES, PRETRIAL
P
UBLICITY, AND THE DUKE LACROSSE CASE
Many of the traits on exhibit in the Duke lacrosse case characterize not only
crime news, but news of the policy process generally. These include
2. See, e.g., BAYDOUN, supra note 1; TAYLOR & JOHNSON, supra note 1; DON YEAGER & MIKE
PRESSLER, IT’S NOT ABOUT THE TRUTH: THE UNTOLD STORY OF THE DUKE LACROSSE CASE AND
THE
LIVES IT SHATTERED (2007).
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
Autumn 2008] RACE TO JUDGMENT 95
overreliance on public officials, overuse of standardized story scripts and
familiar stereotypes, and “pack journalism”—the tendency of reporters from
nominally competitive news organizations to converge on the same framings. In
the case of crime coverage, these media routines can facilitate a pro-prosecution
slant that appears across news coverage when law-enforcement officials are
eager to promote claims of guilt. Perhaps the most important prerequisite to
achieving the journalistic ideal of balance is the requirement of having reliable,
legitimate, and credible sources competing to advance alternative narratives.
Without such contending elite forces working to impose different frames on the
coverage of an issue, media coverage is usually one-sided and arguably unfair to
at least some participants.
3
News of criminal cases often does not meet the
requirement of contending elites working to uphold alternative narratives.
4
Thus, crime news is almost always heavily slanted toward the prosecution for a
structural reason: unlike in coverage of policy debates over, say, social security,
abortion, or the environment, there is no institutionalized political-party system
to provide a more-or-less automatic two-sided debate among credible elites that
journalists can reflect in their coverage.
Pretrial journalism, in particular, tends generally to treat the presumption of
innocence as a formality, largely limited to using the word allege, without
actually covering the story in a balanced fashion that makes clear the possibility
that district attorneys (D.A.s), police, and judges (and juries) can make
mistakes or have bureaucratic, psychological, or political motives.
5
Although relatively few crimes receive sustained media attention,
6
when
they do, coverage is likely to include information that can be considered
prejudicial under American Bar Association (ABA) guidelines. In covering
crime stories, journalists typically rely on law-enforcement officials’ views,
downplaying the defense perspective while minimally acknowledging the
3. W. LANCE BENNETT ET AL., WHEN THE PRESS FAILS: POLITICAL POWER AND THE NEWS
MEDIA FROM IRAQ TO KATRINA (2007) (examining the implications of news framing on, inter alia, the
Abu Ghraib incidents); R
OBERT M. ENTMAN, PROJECTIONS OF POWER: FRAMING NEWS, PUBLIC
OPINION, AND U.S. FOREIGN POLICY (2004) (examining the implications of news framing on public
opinion in media coverage of U.S. foreign policy) [hereinafter P
ROJECTIONS]; Robert M. Entman,
Framing Bias: Media in the Distribution of Power, 57 J. C
OMM. 163, 164–70 (2007) [hereinafter Framing
Bias] (discussing the way in which framing and biases in media coverage interact in the creation of news
slant); Robert M. Entman et al., Doomed to Repeat: Iraq War News, 2003-2007, 52 A
M. BEHAV.
SCIENTIST (Dec. 2008) (suggesting that a key factor facilitating the one-sided framing of U.S. foreign
policy is the fact that “the magnitude and cultural resonance of media attention to negative results of
foreign policy tends to go down as they become more commonplace” (emphasis in original)).
4. See P
ROJECTIONS, supra note 3, at 151.
5. This was certainly an issue in the Duke lacrosse case, when the prosecutor’s potential ulterior
motives were perhaps more obvious than normal. Although journalists paid some attention to the
connection between the case and District Attorney Michael Nifong’s upcoming election campaign, in
hindsight it seems they gave his claims undue deference, perhaps because it seemed literally incredible
that a D.A. in the national-media glare would engage in blatant misconduct.
6. Ralph Frasca, Estimating the Occurrence of Trials Prejudiced by Press Coverage, 72
J
UDICATURE 162, 165 (1988).
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
96 LAW AND CONTEMPORARY PROBLEMS [Vol. 71:93
innocence presumption.
7
Thus, news of crime generally exhibits a pro-
prosecution bias, rooted most importantly in this dependence of reporters on
official and, therefore, purportedly credible sources.
8
One study measuring the
extent of pretrial publicity on Los Angeles television news found that nineteen
percent of the defendants in crime stories were associated with at least one
category of potentially prejudicial information as defined by the ABA.
9
Another study found that twenty-seven percent of suspects in crime stories
were described using prejudicial information.
10
Most of this information was
cited to law-enforcement officers and prosecutors.
11
The imbalanced perspective in the news, facilitated by overreliance on
public officials, is compounded by the inability of criminal-defense teams, even
well-funded ones, to counter the public-relations machinery of law-enforcement
agencies. The latter’s spokespersons enjoy established relationships with local
and wire-service reporters, and they serve superiors with strong career
incentives to maximize publicity for crime-fighting successes.
12
Beyond this, few
reporters—especially those from national news organizations parachuting into a
local scene like Durham—have the time to investigate documentary evidence,
to interview ordinary citizens, or to otherwise probe alternative sources of
information. Defendants face further jeopardy from the very facts that a district
attorney has indicted and police officers have arrested them. For most people,
these actions connote guilt because they believe (based perhaps in part on
television crime shows like Law and Order and CSI) that officials take these
7. See Michael Welch, Melissa Fenwick & Meredith Roberts, Primary Definitions of Crime and
Moral Panic: A Content Analysis of Experts’ Quotes in Feature Newspaper Articles on Crime, 34 J.
RES.
CRIME & DELINQ. 474, 474, 488 (1997) (demonstrating a general tendency in newspaper crime
coverage to rely on law-enforcement officials and to underrepresent the perspectives of defense
lawyers).
8. See id. The defense attorneys, the contending side, are likely viewed as more biased than
prosecutors—their job is to try to get their clients off—and thus their claims are likely to be viewed
more skeptically.
9. Travis L. Dixon & Daniel Linz, Television News, Prejudicial Pretrial Publicity, and the
Depiction of Race, 46 J. B
ROADCASTING & ELECTRONIC MEDIA 112, 127–29 (2002) (reporting
findings based on a sample that yielded two seven-day composite weeks of news programming with 200
programs in all).
10. Dorothy J. Imrich et al., Measuring the Extent of Prejudicial Pretrial Publicity in Major
American Newspapers: A Content Analysis, 45 J. C
OMM. 94, 94 (1995).
11. Id.; see also Celeste Michelle Condit & J. Ann Seltzer, The Rhetoric of Objectivity in the
Newspaper Coverage of a Murder Trial, 2 C
RITICAL STUD. MASS COMM. 197, 197 (1985) (showing
subtle pro-prosecution bias in newspaper reporting on a criminal trial); Nancy Mehrkens Steblay et al.,
The Effects of Pretrial Publicity on Juror Verdicts: A Meta-Analytic Review, 23 L
AW & HUM. BEHAV.
219, 219 (1999) (suggesting that negative pretrial publicity typically and significantly promotes guilty
judgments); Christina A. Studebaker & Steven D. Penrod, Pretrial Publicity: The Media, the Law, and
Common Sense, 3 P
SYCHOL. PUB. POL’Y & L. 428 passim (1997) (using the term “pretrial publicity” as
synonymous with prejudicial pretrial publicity, suggesting that legal scholars and many lawyers presume
as a matter of course that publicity will negatively slant against defendants).
12. See Edward Stringham, Justice After Nifong, W
ASH. TIMES, May 22, 2007, at A15 (“[T]he
incentives created by our highly politicized legal system . . . reward[] law enforcement officials for high
conviction rates, rather than meting out justice. Indeed, in high-profile cases, law enforcement officials
frequently use a conviction to advance their career.”).
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
Autumn 2008] RACE TO JUDGMENT 97
actions only after amassing strong evidence.
13
Journalists no less than their
audiences can fall prey to these not unreasonable assumptions.
III
R
ESEARCH ON BLACKS, WHITES, CRIME, AND NEWS
This section sets a context for evaluating coverage of the Duke lacrosse case
by considering the normal in crime news: a largely unintentional slant that
reinforces whites’ antagonism, particularly toward black defendants, and more
generally equates blacks with crime and danger. As this section shows, virtually
all serious academic research reveals a consistent bias against African
Americans, both in the coverage of crime and more generally.
14
At the same
time, media content—not just news but entertainment and “infotainment”—
usually promotes white privilege and the idea that whites occupy the top of a
racial hierarchy wherein blacks are largely and naturally relegated to the
bottom.
15
Thus the Duke lacrosse case is the exception that proves the rule, an
inversion of the normal situation in which black defendants in criminal cases are
especially likely to be presumed guilty because they are subject to the
stereotypes or heuristics that most whites apply to the category “black
person.”
16
A. The Racial Nature of Crime Coverage
Media stereotypes consist of recurring messages that associate persons of
color with traits, behaviors, and values generally considered undesirable,
inferior, or dangerous. In the context of crime coverage, there is considerable
evidence that media portray blacks and Latinos as criminal and violent.
17
These
images matter because they are a central component in a circular process by
which racial and ethnic misunderstanding and antagonism are reproduced, and
13. See, e.g., RONALD WEITZER & STEVEN A. TUCH, RACE AND POLICING IN AMERICA:
CONFLICT AND REFORM 41 (2006) (reporting that 86% of whites, 73% of African Americans, and 80%
of Latinos were either “Very” or “Somewhat Satisfied” with the police department in their city and
very few were “Very Dissatisfied”); Lawrence D. Bobo & Victor Thompson, Unfair by Design: The
War on Drugs, Race, and the Legitimacy of the Criminal Justice System, 73 S
OC. RES. 445, 454-62
(showing that whites’ perceptions of the fairness of the criminal-justice system exceeds blacks’
perceptions on multiple measures).
14. On media and race generally, see R
OBERT M. ENTMAN & ANDREW ROJECKI, THE BLACK
IMAGE IN THE WHITE MIND: MEDIA AND RACE IN AMERICA (2000).
15. See generally R
OBERT ENTMAN, YOUNG MEN OF COLOR IN THE MEDIA: IMAGES AND
IMPACTS, JOINT CENTER FOR POLITICAL AND ECONOMIC STUDIES BACKGROUND PAPER (2006).
16. See generally Patricia G. Devine, Stereotypes and Prejudice: Their Automatic and Controlled
Components, 56 J.
PERSONALITY & SOC. PSYCHOL. 5, 5 (1989) (demonstrating that both “high- and
low-prejudiced persons are . . . knowledgeable of the cultural stereotype [of blacks]. . . . [This]
stereotype is automatically activated . . . [and] low-prejudice responses require controlled inhibition of
[this] automatically activated stereotype.”). Thus, when individuals are not consciously monitoring their
stereotypes, both the prejudiced and the unprejudiced produce stereotype-consistent evaluations. See
id. Blacks are often stereotyped as lazy and violent, contributing further to presumptions of criminal
tendencies. P
AUL M. SNIDERMAN & THOMAS PIAZZA, THE SCAR OF RACE 38–46 (1993).
17. See infra notes 18–39 and accompanying text.
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
98 LAW AND CONTEMPORARY PROBLEMS [Vol. 71:93
thus become predictable influences in the criminal-justice process. Such
coverage may reinforce biases against black and Latino defendants even in the
absence of specific pretrial publicity about a given defendant’s case. Here are
some examples from the extensive research literature detailing the varied and
subtle ways in which television and print news stereotype blacks and Latinos in
crime coverage:
1. Blacks and Latinos are more likely than whites to appear as lawbreakers
in news—particularly when the news is focusing on violent crime.
18
Blacks and Latinos are more likely to appear as perpetrators than
victims.
19
Blacks are overrepresented as perpetrators of violent crime
when news coverage is compared with arrest rates.
20
Work that engages
in “inter-reality” comparisons suggests the media portrayal is somewhat
less distorted than the comparison of the number of black, white, and
18. For examples of studies reaching this conclusion through “intergroup comparisons” (that is,
comparing the extent to which members of different racial groups are portrayed as perpetrators) based
on evidence from systematic content analyses, see Travis L. Dixon & Daniel Linz, Overrepresentation
and Underrepresentation of African Americans and Latinos as Lawbreakers on Television News, 50 J.
C
OMM., Spring 2000 at 131, 149 (local television news programming in Los Angeles and Orange
counties, Cal.); E
NTMAN & ROJECKI, supra note 14 (multiple media formats and locations); Daniel
Romer et al., The Treatment of Persons of Color in Local Television News: Ethnic Blame Discourse or
Realistic Group Conflict?, 25 C
OMM. RES. 286, 294–98 (1998) (local television news programming in
Phila., Pa.). See also Travis L. Dixon & Cristina L. Azocar, The Representation of Juvenile Offenders by
Race on Los Angeles Area Television News, 17 H
OW. J. COMM. 143, 143, 151 (2006) (portrayal of
juvenile lawbreakers in local television news programming in L.A., Cal.).
19. For examples of studies employing “inter-role comparisons” (that is, comparing the extent to
which members of a specific racial or ethnic group are portrayed in more- and less-sympathetic roles)
to find that people of color are more likely to be portrayed as criminals than as victims, see Travis L.
Dixon & Daniel Linz, Race and the Misrepresentation of Victimization on Local Television News, 27
C
OMM. RES. 547, at 559–560 (2000); Romer et al., supra note 18, at 294–98, (finding also that blacks are
more likely to be portrayed as perpetrators than as bystanders, experts, or other nonperpetrators).
20. For examples of studies reaching this conclusion through “inter-reality comparisons” (that is,
comparing media portrayals with “reality” by comparing content analyses to arrest statistics), see
Dixon & Linz, supra note 18, at 131 (finding blacks, though not Latinos, overrepresented as
lawbreakers); Franklin D. Gilliam, Jr. et al., Crime in Black and White: The Violent, Scary World of
Local News, 1 H
ARV. INT’L. J. PRESS/POL., June 1, 1996, at 6, 10–12 (finding blacks, though not
Latinos, overrepresented as lawbreakers in coverage of violent and nonviolent crimes). See also Dixon
& Azocar, supra note 18, at 143, 152–53 (portraying juvenile lawbreakers). But see Franklin D. Gilliam,
Jr. & Shanto Iyengar, Prime Suspects: The Influence of Local Television News on the Viewing Public, 44
A
M J. POL. SCI. 560, 562 (2000) (finding that blacks appear in crime coverage at rates not much
different than the actual black arrest rate for Los Angeles County, Cal.); Ted Chiricos & Sarah
Eschholz, The Racial and Ethnic Typification of Crime and the Criminal Typification of Race and
Ethnicity in Local Television News, 39 J. R
ES. CRIME & DELINQ. 400, 400 (2002) (finding blacks
underrepresented when comparing the racial composition of television news suspects on Orlando, Fla.
TV news with the percentages of arrests involving blacks and whites). Of course, in this context, the
heavy emphasis on violent crime in television news coverage becomes especially troubling. See
generally L
ORI DORFMAN & VINCENT SCHIRALDI, OFF BALANCE: YOUTH, RACE, AND CRIME IN THE
NEWS (2001) (surveying the representation of crime in the news media in connection with youth and
race). There is no natural law that makes street crime and violence automatically more newsworthy
than city-budget debates, or education, mass transit, and other policy discussions to which local TV
news typically pays comparatively little attention. Summarizing the results of these content analyses is
somewhat complicated by different standards of comparison (for example, comparing the extent to
which members of different racial groups are portrayed as perpetrators versus comparing the media
portrayal with “reality” through the use of measures such as arrest statistics). See, e.g., Dixon & Linz,
supra note 18, at 131 (employing intergroup and inter-reality comparisons).
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
Autumn 2008] RACE TO JUDGMENT 99
Latino perpetrators might imply.
21
Nonetheless, the picture is troubling,
suggesting that news cultivates the perception of blacks and Latinos as
lawbreakers.
2. In contrast, whites are overrepresented as victims of violence and as
law-enforcers, while blacks are underrepresented in these sympathetic
roles.
22
3. These patterns are more disturbing when one considers that blacks and
Latinos may not appear as frequently as whites outside the crime-news
context. Because of the values that define the newsworthy, blacks in
criminal roles tend to outnumber blacks in socially positive roles in
newscasts and daily newspapers.
23
In some areas, at least, Latinos may
receive even worse treatment than blacks. Chiricos and Eschholz
examined three weeks of local-news programming in Orlando and found
that one in twenty whites appearing on the local television news was a
criminal suspect compared with one in eight blacks and one in four
Latinos.
24
4. Studies of local news in Chicago and elsewhere suggest that depictions
of black suspects (mostly young men) tend to be more symbolically
21. Measures of reality to which representations are compared are deeply problematic. Assume
blacks and Latinos are represented in proportion to their actual crime rates according to some
relatively simple measures (for example, blacks commit 25% of crimes in a metropolitan area and
constitute 25% of accused criminals in local news reports). This leaves aside many ambiguities in the
guidelines offered by proportionality as a standard: for example, does accuracy demand that today’s
coverage offer racial proportionality in terms of last week’s racial crime rate, or this year’s, or last
year’s, or should it reflect rate of change in relative racial crime rates? If blacks committed 35% of
armed robberies last month, what happens if they commit 12% this month, which one finds out only
when the statistics are compiled at some later date? Should one underreport the following month’s
black-committed armed robberies to make up for the previous month’s overrepresentation? Exactly
how should coverage be calibrated to severity of crime committed by members of racial groups—by
devoting more time, more-vivid language, more details on the victims? And how does one measure
“severity” anyway? For further discussion, see E
NTMAN & ROJECKI, supra note 18 at 213–15. The list
could go on. More importantly, even if blacks and Latinos are somehow represented in “proper”
proportion, a lack of contextual reporting reinforces whites’ ignorance and tendencies to stereotype. As
an example of needed context, black crime rates among young adult males are no different from white
rates when controlling for employment status. In other words, unemployment is a primary cause of
black crime, unemployment rooted in discrimination, poor education systems, and other structural
causes. See W
ILLIAM JULIUS WILSON, WHEN WORK DISAPPEARS: THE WORLD OF THE NEW URBAN
POOR 22 (1996).
22. Travis L. Dixon et al., The Portrayal of Race and Crime on Television Network News, 47 J.
BROADCASTING & ELECTRONIC MEDIA 498, 498 (2003); see also Chiricos & Eschholz, supra note 20,
at 20 (finding “[b]lacks . . . more likely to appear as criminal suspects than as victims or positive role
models”); Dixon & Linz, supra note 18, at 131 (finding whites overrepresented as law defenders on
local news programming in L.A., Cal., though blacks are neither over- nor underrepresented compared
to county employment rates); Dixon & Linz, supra note 19, at 564 (finding, via intergroup comparisons,
whites more likely than blacks or Latinos to be portrayed as victims of crime, and whites
overrepresented as homicide victims on local television news); Romer et al., supra note 18, at 286
(finding “Whites overrepresented as victims of violence compared to their roles as perpetrators, and
persons of color . . . overrepresented as perpetrators of violence against White[s]”).
23. See E
NTMAN & ROJECKI, supra note 18, at 86.
24. Chiricos & Eschholz, supra note 20, at 407, 416. Though this same study did not find that blacks
were overrepresented when looking only among TV-news suspects, id. at 410, this lack of other
positive-role portrayals for blacks reinforces the link between blacks and crime.
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
100 LAW AND CONTEMPORARY PROBLEMS [Vol. 71:93
threatening than those of whites accused of similar crimes.
25
Black
defendants in one study were more likely to be shown in mug shots.
26
In
the ubiquitous “perp walks,” blacks were twice as likely as whites to be
shown under some form of physical restraint by police—although all
were accused of scary and generally violent crimes.
27
In some notorious,
highly publicized crimes—such as the 1989 alleged rape of a wealthy,
young white woman in Central Park by a “gang” of Latino and black
young men—young men of color appear particularly susceptible to
portrayals that associate them with extreme threat and less-than-human
traits. Narratives routinely used such words as “savage” and “wild.”
28
5. Violence and youth, especially male youth, are closely linked: most
stories that feature young people on local news depict violence they
commit or suffer, and in those stories, older white men are the dominant
speakers.
29
Local news does not often portray young persons as
positively contributing to society.
30
6. People of color are more likely to be subjected to negative pretrial
publicity. One study found that black and Latino defendants are twice as
likely as white defendants to be subjected to negative pretrial publicity
and that defendants who victimized whites were more likely to have
prejudicial information broadcast about them than defendants who
victimized nonwhites.
31
25. See, e.g., ENTMAN & ROJECKI, supra note 18, at 78–93 (reporting and discussing the results of a
study on the use of racial stereotypes in local-crime news coverage in Chicago, Ill.); Chiricos &
Eschholz, supra note 20, at 20 (finding that blacks, and especially Latinos, who appeared as suspects in
Orlando, Fla., local-crime news coverage appeared in more threatening contexts than whites).
26. E
NTMAN & ROJECKI, supra note 18, at 82.
27. Id. at 83–84. One complexity to note is that it may well be true that police officers physically
restrain more black defendants than they restrain white defendants because of the officers’ own racial
fears. News reports that may cultivate negative stereotyping and fear of young men of color may also
accurately convey aspects of the “real world.” This consideration indicates that solving the problem
may require more than simply urging journalists to report on discrete events or specific facts more
accurately.
28. See generally Michael Welch et al., Moral Panic Over Youth Violence: Wilding and the
Manufacture of Menace in the Media, 34 Y
OUTH & SOC’Y 3 (2002) (discussing use of the term “wilding”
and synonyms for “wild” in media coverage following this incident). Later evidence showed that the
idea of a gang of nonwhites repeatedly assaulting the woman was exaggerated and most of the accused
were exonerated. See Jim Dwyer, New Slant on Jogger Case Lacks Official Certainty, N.Y.
TIMES, Jan.
28, 2003, at B7; Susan Saulny, Convictions and Charges Voided In ‘89 Central Park Jogger Attack, N.Y.
TIMES, Dec. 20, 2002, at A1.
29. Lori Dorfman & Katie Woodruff, The Roles of Speakers in Local Television News Stories on
Youth and Violence, 26 J.
POPULAR FILM & TELEVISION 80, 81, 83 (1998).
30. See id.; D
ORFMAN & SCHIRALDI, supra note 20. This coverage leads to specific perceptions of
juvenile offenders. See Robert K. Goidel et al., The Impact of Television Viewing on Perceptions of
Juvenile Crime, 50 J.
BROADCASTING & ELECTRONIC MEDIA 119, 134 (2006) (reporting the results of
a survey of Louisiana residents that found viewers of television news were more likely to believe
juvenile crime was increasing, more likely to overestimate the percentage of juvenile offenders
imprisoned for violent crimes, and more likely to believe rehabilitation in prison was more effective
than community-based rehabilitation programs).
31. Dixon & Linz, supra note 9, at 128–29 (finding also that, accounting for the race of the victim,
being black with a white victim more than doubles the odds of prejudicial pretrial information
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
Autumn 2008] RACE TO JUDGMENT 101
7. Several studies show that black victims are less likely to be covered than
white victims in newspaper coverage of crime.
32
But unlike studies
focused on local television, studies of newspapers do not indicate that
minorities are overrepresented as perpetrators.
33
Newspaper coverage is
somewhat better than local television news on some dimensions, if only
because it does not follow the “if it bleeds it leads” norm and devotes
much less of its news space to crime.
34
Stories themselves are often
longer and, on average, perhaps more likely to contain more context
than TV. On the other hand, the additional words devoted to crime by
newspapers might yield a greater net volume of negative pretrial
information.
35
8. Aside from crime, perhaps the most frequent and disproportionate
association made with persons of color in the news media is poverty. In
stories featuring poverty as a topic, newsmagazines like Time and
Newsweek overrepresent blacks, while underrepresenting whites,
Latinos, and Asians.
36
At the same time, the magazines overrepresent
appearing; a Latino with a white victim more than triples the odds that prejudicial information will be
aired when compared with a white defendant and white victim).
32. See, e.g., John W.C. Johnstone et al., Homicide Reporting in Chicago Dailies, 71 J
OURNALISM
Q. 860, 860 (1994) (homicide coverage in Chicago, Ill., newspapers); David Pritchard & Karen D.
Hughes, Patterns of Deviance in Crime News, 47 J. C
OMM., Summer 1997, at 49, 49 (homicide coverage
in Milwaukee, Wis., newspapers); Susan B. Sorenson et al., News Media Coverage and the
Epidemiology of Homicide, 88 A
M. J. PUB. HEALTH 1510, 1510 (1998) (comparing a special, seven-part
series on homicide in the Los Angeles Times to the homicides that occurred during the period featured
in the report and finding that homicides of blacks and Latinos were substantially underreported);
Alexander Weiss & Steven Chermak, The News Value of African-American Victims: An Examination
of the Media’s Presentation of Homicide, 21 J. C
RIME & JUST. 71, 71 (1998) (homicide coverage in
Indianapolis, Ind., newspapers).
33. See, e.g., David Pritchard, Race, Homicide, and Newspapers, 62 J
OURNALISM Q. 500, 500
(1985) (finding homicides with minority suspects get less coverage in Milwaukee newspapers); Shelly
Rogers et al., ‘Reality’ in the St. Louis Post-Dispatch, 21 N
EWSPAPER RES. J., Summer 2000, at 51, 63 (
“African Americans did not dominate the stereotyped role[] of perpetrator” in St. Louis, Mo.,
newspaper coverage.); Susan Sorenson et al., supra note 32, at 1511 (finding suspect characteristics like
race unrelated to coverage in the Los Angeles Times with one exception of interest here: “Latino
suspects were less likely than others to receive coverage . . . .”). But see Weiss & Chermak, supra note
32, at 78 (finding that, although the number of articles in coverage of murder in Indianapolis
newspapers was similar whether the suspect was black or white, the average length of the article was
longer if the suspect was black); Tara-Nicholle Beasley Delouth & Cindy J.P. Woods, Biases Against
Minorities in Newspaper Reports of Crime, 79 P
SYCHOL. REP. 545 (1996) (finding some evidence, with
admittedly small numbers necessitating caution about the conclusions, that disclosure of the victims’
identity was more common when the suspect was an ethnic minority).
34. Newspapers emphasize violence less in their news than television. See E
NTMAN & ROJECKI,
supra note 18, at 88–90 (comparing newspaper and local television-news coverage); Joseph F. Sheley &
Cindy D. Ashkins, Crime, Crime News, and Crime Views, 45 P
UB. OPINION Q. 492, 492 (1981)
(comparing newspaper and television-news coverage with police figures). Crime is the dominant topic
on local television news and an important topic in metropolitan daily papers. See R
ICK EDMONDS &
THE PROJECT FOR EXCELLENCE IN JOURNALISM, THE STATE OF THE NEWS MEDIA 2006,
http://www.stateofthenewsmedia.com/2006/narrative_newspapers_contentanalysis.asp?cat=2&media=3
(reporting results of a content analysis of news coverage by local and national television news and
newspapers on May 11–12, 2006).
35. Dixon & Linz, supra note 9, at 130–31.
36. See Rosalee Clawson & Rakuya Trice, Poverty as We Know It: Media Portrayals of the Poor, 64
P
UB. OPINION Q. 53, 56–57 (2000) (analyzing media depictions of poverty demographics).
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
102 LAW AND CONTEMPORARY PROBLEMS [Vol. 71:93
the nonworking poor and the urban poor in their illustrations of
poverty.
37
In this way, not only do media encode poverty as an especially
black trait, but they undermine potential sympathy, especially among
the white majority, for antipoverty programs.
38
Poverty can be portrayed
as a condition that merits sympathy, but it more often appears
associated with threats in the form of crime, violence, drugs, gangs, and
aimless activity.
39
Such messages not only stereotype the blacks and Latinos who are featured
in them, but also contribute to a stereotypical association between blacks,
criminality, and guilt that can influence evaluations and behavior. These
messages also reinforce negative emotions and a sense of social distance that
may promote a belief in inherent group conflict between blacks or Latinos and
whites. Moreover, these stereotypes arise not merely from the news, but from
TV and film entertainment, advertising, and sports programming as well.
40
The implications of this research for public attitudes are troubling. Messages
continually associating people of color, especially blacks, with poverty and
crime reinforce the updated form of racial prejudice known as symbolic racism,
racial resentment, or racial animosity.
41
In the absence of contextual information
about discrimination and other forces that produce the poverty, racialized
images of poverty reinforce the stereotype that blacks are lazy and therefore
deserve their impecuniousness. Racialized crime coverage reinforces the
37. See id. at 60.
38. See generally E
NTMAN & ROJECKI, supra note 18, at 103–06 (discussing the implications of
news-media portrayals of race and poverty on, inter alia, racial animosity); M
ARTIN GILENS, WHY
AMERICANS HATE WELFARE: RACE, MEDIA, AND THE POLITICS OF ANTIPOVERTY POLICY, (1999)
(examining media portrayals of and attitudes toward race, poverty, and welfare); Robert M. Entman,
Television, Democratic Theory, and the Visual Construction of Poverty, in 7 R
ESEARCH IN POLITICAL
S
OCIOLOGY 139 (Philo C. Washburn ed., 1995) (discussing the effects of race and poverty in news
media on the nonpoor’s support for antipoverty policy); Martin Gilens, “Race Coding” and White
Opposition to Welfare, 90 A
M. POL. SCI. REV. 593 (1996) (“[R]acial attitudes are the single most
important influence on whites’ welfare views.”).
39. See E
NTMAN & ROJECKI, supra note 18, at 96–100 (reporting results of a study of the depiction
of poverty in local newspapers and local and national television news).
40. See R
OBERT M. ENTMAN, YOUNG MEN OF COLOR IN THE MEDIA: IMAGES AND IMPACTS,
15–17 (2006) (discussing the prevalence and implications of racial stereotyping in these contexts). Such
media routines are central to a self-perpetuating cycle that reproduces white racial antagonism directed
at persons of color. The cultural reproduction of racial antagonism among whites as a result of media
coverage reinforces discrimination, which promotes joblessness and hopelessness among African
Americans, and that in turn encourages criminal behavior. This cycle yields continuing crime news that
reinforces the stereotypes and the automatic and unconscious anxieties and associations of blacks with
danger and lawlessness. Figure 1 in Entman offers a graphic representation of this circular process,
which in admittedly oversimplified and abstract form depicts the complex interrelationships among
white elites and the institutions that they dominate; ordinary white citizens’ sentiments, decisions, and
behaviors; and the lives and life chances of persons of color. Id. at 9. See generally E
NTMAN &
ROJECKI, supra note 18.
41. See generally E
NTMAN & ROJECKI, supra note 18 (discussing the interactions of portrayals of
race in media, racial animosity, and symbolic racism); D
ONALD R. KINDER & LYNN M. SANDERS,
DIVIDED BY COLOR: RACIAL POLITICS AND DEMOCRATIC IDEALS (1996) (discussing racial
resentment in the context of U.S. politics); Christopher Tarman & David O. Sears, The
Conceptualization and Measurement of Symbolic Racism, 67 J. P
OL. 731 (2005) (studying and discussing
the effects of symbolic racism on racial policy preferences).
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
Autumn 2008] RACE TO JUDGMENT 103
stereotype that blacks are not just lazy, but violent.
42
Both of these stereotypes
are important components of this new form of racial prejudice. Moreover,
empirical evidence demonstrates associations between racial resentment and
whites’ support of punitive crime policies
43
and opposition to preventative
policies.
44
B. Effects on White Attitudes
The effects on whites of this heavy representation of blacks and Latinos in
crime news have been documented through empirical studies. As Gilliam and
Iyengar argue, coverage of this nature creates a “crime script” in which crime is
violent and perpetrators are black. Repeated exposure to this script promotes
and reinforces negative racial stereotypes.
45
Kang uses the metaphor of the
“Trojan Horse Virus” to describe how local television news of this nature can,
without viewers’ awareness and without intent on the part of news producers,
create and reinforce associations between blacks and violence in the minds of
citizens.
46
These stereotype-based cognitive and emotional responses are often
automatically, quickly, and unconsciously triggered, and go on to affect a wide
range of sentiments.
47
For example, one researcher found that an act (an
ambiguous shove) was interpreted as more violent when performed by a black
than when performed by a white,
48
consistent with the notion that “black” and
“violent” have become linked in the minds of many individuals. Other work
illustrates that when presented with an ambiguous suspect, individuals are more
likely to assume the suspect is black.
49
Further evidence comes from studies
42. See the 2000 General Social Survey. Individuals were asked to rate blacks, Asians, Latinos, and
whites as a group on two traits of relevance here: on a seven-point scale from lazy to hardworking and
on a seven-point scale from violence-prone to not-violence-prone. Thirty-five percent of respondents
rated blacks as lazy (that is, they gave blacks a rating corresponding to the three numbers at the
negative end of the scale for lazy), 21% rated Latinos as lazy, and only 11% rated whites or Asians as
lazy. The stereotypes are more extreme in the case of violence: 48% rated blacks as prone to violence
(that is, they gave blacks a rating that corresponded to the three numbers at the negative end of the
seven-point scale), 38% rated Latinos as prone to violence, 21% rated whites as prone to violence, and
17% rated Asians as prone to violence.
43. See Gilliam & Iyengar, supra note 20, at 560, 571 (“[E]xposure to the racial element of the
[experiment’s] crime script increases [whites’] support for punitive approaches to crime . . . .”).
44. Eva G.T. Green et al., Symbolic Racism and Whites’ Attitudes Towards Punitive and
Preventative Crime Policies, 30 L
AW & HUM. BEHAV. 435, 435 (2006).
45. See Gilliam & Iyengar, supra note 20, at 560–61.
46. See Jerry Kang, Trojan Horses of Race, 118 H
ARV. L. REV. 1489, 1553–54, 1562–63 (2005).
Kang draws on psychological work on implicit attitudes, id. at 1506–14 (citing, for example, Brian A.
Nosek, Mahzarin Banaji & Anthony G. Greenwald, Harvesting Implicit Group Attitudes and Beliefs
from a Demonstration Web Site, 6 G
ROUP DYNAMICS 101, 105 (2002)), in describing how coverage may
be detrimental. Implicit-attitudes research shows that whites express some explicit preference for their
own group but much greater levels of implicit preference. Id. at 1513. Even individuals who honestly
self-report positive attitudes toward individuals in other racial categories may hold implicit negative
attitudes toward that group. Id.
47. Id. at 1515–48.
48. B.L. Duncan, Differential Social Perception and Attribution of Intergroup Violence: Testing the
Lower Limits of Stereotyping of Blacks, 34 J. P
ERSONALITY & SOC. PSYCHOL. 590, 590 (1976).
49. See Gilliam & Iyengar, supra note 20, at 564 (finding that subjects who viewed a crime story
with no mug shot were much more likely to misremember seeing a black perpetrator, consistent with
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
104 LAW AND CONTEMPORARY PROBLEMS [Vol. 71:93
using a computerized experiment in which participants were instructed to shoot
an armed target but not an unarmed target.
50
The target individual was either
black or white, was holding either a gun or another object, and was presented in
a realistic background.
51
Under time pressure, participants were more likely to
mistakenly shoot an unarmed black target than an unarmed white target and
were more likely to mistakenly fail to shoot an armed white target than an
armed black target.
52
In other words, participants in the study were more likely
to assume wrongly that a black individual was armed and assume wrongly that a
white individual was not.
53
Other research reveals the interaction between these
experimental results and media images.
54
This work suggests that the shooting
bias results from the mental linkage of black males to guns—a connection
cemented by vivid media images—rather than from generalized racial prejudice
against blacks.
55
With this crime–guns–blacks schema so prevalent, it is not
the claim that crime news reinforces a script in which crime is violent and perpetrators are black); see
also Mary Beth Oliver, Caucasian Viewers’ Memory of Black and White Criminal Suspects in the News,
49 J. C
OMM., Summer 1999, at 46, 46 (finding that white viewers presented with a newscast featuring a
white murder suspect were increasingly likely over time to misidentify the suspect as black); Mary Beth
Oliver & Dana Fonash, Race and Crime in the News: Whites’ Identification and Misidentification of
Violent and Nonviolent Criminal Suspects, 4 M
EDIA PSYCHOL. 137, 137 (2002) (finding that white
viewers presented with a newspaper crime brief featuring violent- and nonviolent-crime stories and
photographs of white and black criminal suspects had an increased likelihood later to misidentify the
black criminal suspects in the violent-crime stories); Travis L. Dixon, Black Criminals and White
Officers: The Effects of Racially Misrepresenting Law Breakers and Law Defenders on Television News,
10 M
EDIA PSYCHOL. 270, 270 (2007) (finding that test subjects were highly likely to assume that
unidentified perpetrators were black no matter what their prior level of news viewing and that heavy
viewers of news were more likely to say an unidentified police officer was white). Dixon’s finding that
unidentified police officers were more likely to be identified as white is consistent with the notion that
television viewing creates specific crime scripts. See id. Interestingly, Dixon’s finding that news viewing
did not moderate the likelihood of rating the unidentified suspect as black suggests that this link has
been made for all viewers and is chronically accessible. See id. at 281.
50. Joshua Correll et al., The Police Officer’s Dilemma: Using Ethnicity to Disambiguate Potentially
Threatening Individuals, 83 J. P
ERSONALITY & SOCIAL PSYCHOL. 1314, 1314 (2002) [hereinafter
Correll et al., Police Officer’s Dilemma]; see also Joshua Correll et al., The Influence of Stereotypes on
Decisions to Shoot, 37 E
UR. J. SOC. PSYCHOL. 1102, 1102, 1105 (2007) [hereinafter Correll et al.,
Decisions to Shoot] (replicating this study and adding a preliminary task in which subjects were asked
to read news stories about black or white criminals).
51. Correll et al., Police Officer’s Dilemma, supra note 50, at 1315–17.
52. Id. at 1319.
53. See id.; Correll et al., Decisions to Shoot, supra note 50, at 1102 (finding exposure to news
stories featuring black criminals increased shooting bias); see also Kang, supra note 46, at 1562–63
(describing how local television news can create associations between racial categories and violent
crime).
54. See B. Keith Payne, Prejudice and Perception: The Role of Automatic and Controlled Process in
Misperceiving a Weapon, 81 J. P
ERSONALITY & SOCIAL PSYCHOL. 181, 181 (2001) (finding participants
identified guns faster and misidentified tools as guns more often when first presented with photographs
of black faces than when first presented with photographs of white faces).
55. See id. It is worth noting that the shooting experiments show that black subjects, not just
whites, respond prejudicially—suggesting the depth to which this particular cultural equation
penetrates. See Correll et al., Police Officer’s Dilemma, supra note 50, at 1325; Correll et al., Decisions
to Shoot, supra note 50, at 1108–11. But see Kristin A. Lane et al., Implicit Social Cognition and Law, 3
A
NN. REV. L. & SOC. SCI. 427, 435 (2007) (reporting that the same is not true of the Implicit-Attitude
Test experiments, in which blacks respond quite differently from whites to subtle racial cues).
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
Autumn 2008] RACE TO JUDGMENT 105
surprising that there are consequences for how people think about politics and
public policy.
Researchers have probed the effects of media representations of race and
crime on whites’ fearfulness of crime and whites’ tendencies to support punitive
public policies, such as capital punishment and mandatory long sentences.
Empirical studies have explored the effects of overrepresentation of black
lawbreakers. Using both experimental and survey studies, researchers have
shown that exposure to images of black, male, criminal defendants increased
whites’ punitive attitudes toward crime, as well as their tendencies to endorse
dispositional explanations for criminal behavior and racist beliefs.
56
Dixon and
Azocar draw similar conclusions, finding that news coverage interacts with
“news use” (the amount of news-watching engaged in by TV viewers) to
influence judgments of the structural limitations that blacks face in the United
States, attitudes toward the death penalty, and views on culpability.
57
Other
work reveals conditional effects of exposure to racialized crime depending on
the racial mix of the neighborhoods in which individuals lived. Whites who lived
in homogeneously white neighborhoods endorsed punitive crime policy and
expressed more-negative stereotypes of blacks when exposed to a black suspect
in a violent crime story, whereas whites living in heterogeneous neighborhoods
were unaffected or were moved in the opposite direction.
58
Other studies show an interaction between media exposure and the ethnic
composition of a community in influencing fear of crime. In areas where whites
56. See Gilliam & Iyengar, supra note 20, at 560. In this study, people were exposed to a no-crime
story, a murder story with no information identifying the race of the perpetrator, a murder story with a
mug shot of a black perpetrator, or a murder story with a mug shot of a white perpetrator. Id. at 563.
The murder story was identical across conditions with the exception of the mug shot. Id. Whites
exposed to the black perpetrator showed significantly greater support for punitive remedies for crime
than those who did not view a crime story. Id. at 567–68. Exposure to the black-perpetrator and the no-
perpetrator conditions also increased dispositional attributions among whites. Id. There was no similar
effect among those who were exposed to the white perpetrator. Id. Gilliam and Iyengar also found that
all crime stories influence racial attitudes (including the white-mug-shot condition), though the effects
are greatest in the no-perpetrator condition, then the black-perpetrator condition, and then the white-
perpetrator condition. Id. at 567–68. Using data from the Los Angeles County Social Survey, the
researchers demonstrate that frequent viewers of television news (who presumably get this specific
script) are more punitive. Id. at 570–71. The survey results lend external validity to the experimental
results.
57. Travis L. Dixon & Cristina L. Azocar, Priming Crime and Activating Blackness: Understanding
the Psychological Impact of the Overrepresentation of Blacks as Lawbreakers on Television News, 57 J.
C
OMM. 229, 229 (2007) (finding that heavy viewers of television news who are faced with unidentified
(that is, no race given) suspects are less likely to say blacks face structural limits to success and more
likely to support the death penalty than heavy viewers who are exposed to noncrime stories; and that
exposure to crime news with a majority of black suspects leads people to evaluate a race-unidentified
suspect as more culpable, an effect that is enhanced among heavy news viewers); see also Travis L.
Dixon, Psychological Reactions to Crime News Portrayals of Black Criminals: Understanding the
Moderating Roles of Prior News Viewing and Stereotype Endorsement, 73 C
OMM. MONOGRAPHS 162,
162 (2006) (finding that, after exposure to a newscast featuring either a majority of black or
unidentified-race suspects, viewers who endorsed stereotypes of blacks were more likely to support the
death penalty than viewers who rejected stereotypes of blacks).
58. Franklin D. Gilliam, Jr. et al., Where You Live and What You Watch: The Impact of Racial
Proximity and Local Television News on Attitudes About Race and Crime, 55 P
OL. RES. Q. 755, 755,
760, 770. (2002).
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
106 LAW AND CONTEMPORARY PROBLEMS [Vol. 71:93
perceive that significant proportions of their neighbors are black or Latino,
heavy viewers of news, of “reality” crime shows, and of fictional crime shows
are all markedly more fearful of crime.
59
The media images—and other political and cultural forces—that cultivate
racialized fear and animosity influence public policies through these effects on
prejudice. Soss and his colleagues show that whites’ racial sentiments are
strongly related to their level of support for the death penalty.
60
Furthermore,
residential proximity heightens this effect. In the presence of large populations
of African Americans, whites’ racial antagonism is particularly associated with
high support of the death penalty.
61
If such a strong yet simple equation between
violent crime and blacks were not made in the media and in U.S. culture, such a
relationship would not emerge. After all, murder is murder, and capital
punishment does not inherently implicate race. Nonetheless, historical study
reveals clear relationships among support of capital punishment, racial
attitudes, and the racial composition of different states.
62
On the other side of
the ledger, Weitzer and Tuch show that blacks and Latinos are more likely than
whites to support police-reform policies, in part because of their greater
attention to media reports of police misconduct and the greater extent of their
direct experience with it.
63
Media coverage of crime primes racial sentiments, influencing evaluations of
both candidates and policies. For example, Valentino found that subjects
responded to TV news stories that included brief shots of black, Latino, or
Asian criminal suspects by lowering their ratings of President Clinton—even
59. See Sarah Eschholz, Ted Chirico & Marc Gertz, Television and Fear of Crime: Program Types,
Audience Traits, and the Mediating Effect of Perceived Neighborhood Racial Composition, 50 S
OC.
PROBS. 395, 395, (2003) (finding that viewers of local news, “reality,” or fictional crime shows who
perceive that they live in a neighborhood with a high percentage of blacks have an increased fear of
crime); see also Ted Chiricos et al., Perceived Racial and Ethnic Composition of Neighborhood and
Perceived Risk of Crime, 48 S
OC. PROBS. 322, 322 (2001) (finding that the perception that blacks or
Latinos live nearby influences the perceived risk of criminal victimization); Ted Chiricos et al., Fear, TV
News, and the Reality of Crime, 38 C
RIMINOLOGY 755, 755 (2000) (showing that local news is related to
fear of crime independent of measures of “reality”); Sorin Matei et al., Fear and Misperception of Los
Angeles Urban Space: A Spatial-Statistical Study of Communication-Shaped Mental Maps, 28 C
OMM.
RES. 429, 429 (2001) (finding that people in Los Angeles with heavy television-viewing habits have
increased fear of nonwhite and non-Asian populations). But see Kimberly Gross & Sean Aday, The
Scary World in Your Living Room and Neighborhood: Using Local Broadcast News, Neighborhood
Crime Rates, and Personal Experience to Test Agenda Setting and Cultivation, 53 J. C
OMM. 411, 419–21
(2003) (finding that viewing local television news does not cultivate fear of crime once neighborhood
crime rates and personal experience are taken into account).
60. Joe Soss et al., Why Do White Americans Support the Death Penalty? 65 J. P
OL. 397, 397 (2003).
61. Id. at 409.
62. See Eric P. Baumer et al., Explaining Spatial Variation in Support for Capital Punishment: A
Multilevel Analysis, 108 A
M. J. SOC. 844, 844 (2003) (“[R]esidents of areas with higher homicide rates, a
larger proportion of blacks, and a more conservative political climate are significantly more likely to
support the death penalty.”). See generally David Niven, Bolstering an Illusory Majority: The Effects of
the Media’s Portrayal of Death Penalty Support, 83 S
OC. SCI. Q. 671 (2002) (discussing the complexities
of public sentiments toward the death penalty).
63. Ronald Weitzer & Steven A. Tuch, Reforming the Police: Racial Differences in Public Support
for Change, 42 C
RIMINOLOGY 391, 409–12 (2004).
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
Autumn 2008] RACE TO JUDGMENT 107
though the stories did not mention Clinton or his policies.
64
Presumably, the
widespread publicity about Clinton’s sympathies for persons of color made
white subjects think of him when they perceived crime committed by young
men of color. The 1988 Bush presidential campaign took advantage of whites’
tendencies to automatically associate black young men and Latino young men
with crime and fear in its infamous “Willie Horton” and “Revolving Door”
political advertisements.
65
A related study showed that by merely inserting the
words “inner city” in a survey question about spending money on prisons as
opposed to antipoverty programs, whites’ racial thinking was stimulated to the
point that “racial conservatives” became more punitive, favoring prisons over
antipoverty programs.
66
The mental association between the term “inner city”
and threatening persons of color is apparently so strong that visual images (such
as those in George H.W. Bush’s 1988 campaign advertisements) or explicit
racial labels are not even needed to generate an effect on whites.
67
Physiological
evidence, moreover, also supports the notion that negative associations with
black persons, repeated vividly across the mass media, resonate quite deeply in
whites’ neural networks.
68
C. Race and Crime Representations
The research outlined in this subsection suggests that it would be entirely
inappropriate to generalize from the Duke lacrosse case about the media’s
normal depictions of race and crime. Systematic content analyses of media
depictions of crime and race suggest that the Duke lacrosse case—featuring
white, relatively privileged defendants—does not follow the typical pattern. The
evidence on race and crime news is quite strong, about as strong as any social-
science evidence in the field of communication and public opinion. We know of
no credible research suggesting that wealthy white defendants are
systematically subjected to more unfair coverage or pretrial publicity compared
with black defendants. Moreover, the evidence illustrating the effects of such
coverage suggests that patterns in the media place black defendants at a
64. Nicholas A. Valentino, Crime News and the Priming of Racial Attitudes During Evaluations of
the President, 63 P
UB. OPINION Q. 293, 293 (1999); see also Mark Peffley et al., Racial Stereotypes and
Whites’ Political Views of Blacks in the Context of Welfare and Crime, 41 A
M. J. POL. SCI. 30, 30 (1997)
(“Whites holding negative stereotypes are substantially more likely to judge blacks more harshly than
similarly described whites in the areas of welfare and crime policy.”).
65. See generally, K
ATHLEEN HALL JAMIESON, DIRTY POLITICS: DECEPTION, DISTRACTION,
AND
DEMOCRACY 15–42 (1992) (describing advertisements in the context of negative political
campaigning); T
ALI MENDELBERG, THE RACE CARD 134–68 (2001) (discussing the advertisements in
the context of racial stereotypes in political campaigns).
66. John Hurwitz & Mark Peffley, Playing the Race Card in the Post-Willie Horton Era: The Impact
of Racialized Code Words on Support for Punitive Crime Policy, 69 P
UB. OPINION Q. 99, 99, 108
(2005).
67. See id. at 109.
68. See William A. Cunningham et al., Separable Neural Components in the Processing of Black
and White Faces, 15 P
SYCHOL. SCI. 806, 806 (2004) (using brain-imaging technology to reveal that a
subliminal, black stimulus “lights up” the amygdala, the portion of the brain that is particularly
sensitive to fear).
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
108 LAW AND CONTEMPORARY PROBLEMS [Vol. 71:93
disadvantage because of the association built between blacks, crime, and guilt.
Preexisting racial schemas, therefore, worsen the general problem of prejudicial
pretrial publicity for blacks (and perhaps Latinos). Even in the absence of
specific prejudicial publicity about a given defendant’s case, the general nature
of crime news may prejudice potential jurors against minority defendants. By
contrast, white defendants relative to black defendants may benefit from a
greater (though not necessarily high) presumption of possible innocence.
69
IV
A
NALYSIS OF COVERAGE OF THE DUKE LACROSSE CASE
Just how did the media cover this case? Did the media depict the incident as
a metaphor for larger issues of race and class? Did coverage undermine the
presumption of innocence or appear to slant in favor of one side? To what
extent did the various parties receive positive or negative coverage? Was this an
example of egregious “liberal bias” or “political correctness” run amok in
media coverage, as some have suggested? In order to begin to answer these
questions, we conducted a systematic content analysis, in addition to a more
qualitative assessment of coverage. Media coverage is characterized according
to how the media dealt with the case over the full period that it received
attention in the news; distinctions are also made among time periods (early
coverage compared with coverage after the DNA results were released), among
different media outlets (comparing newspapers with television), and among
different sections of newspapers (distinguishing news from editorials and
commentary).
Before turning to the analysis, however, it is helpful to explore why this case
received more attention than an accusation of rape against three college
students might be expected to generate. A kind of “perfect storm” of events
reinforced the attractions of this case for national, not just local, media. What
contributed to this “storm”? District Attorney Michael Nifong publicly
condemned the students and provided a narrative that was irresistible for the
media. The racial element was an important component of the frame the
District Attorney developed, perhaps intentionally, since he was running for
election and needed to maximize his support within Durham’s large black
electorate.
69. It is also true that because police officers, district attorneys, and the citizen pool that supplies
juries are often dominated by whites, blacks may be more likely to be arrested, treated more rigorously
by the criminal-justice system, and convicted. Moreover, given their lower socioeconomic status, they
are likely to have less-sophisticated and less-able legal representation, and less-refined public-relations
strategies designed to soften up potential jury pools and put political pressure on district attorneys,
police, and judges to treat the defendants fairly. See, e.g., Donna Coker, Foreword: Addressing the Real
World of Racial Injustice in the Criminal Justice System, 93 J. C
RIM. L. & CRIMINOLOGY 827 (2003)
(comparing the empirical evidence of criminal-law enforcement to assumptions underlying Fourth
Amendment and Equal Protection clause jurisprudence); see also Jeffrey T. Ulmer et al., Prosecutorial
Discretion and the Imposition of Mandatory Minimum Sentences, 44 J. R
ES. CRIME & DELINQ. 427, 427
(2007) (“find[ing] that Hispanic males are more likely to receive mandatory minimum[] [sentences] and
that Black–White differences in mandatory application increase with county percentage Black”).
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
Autumn 2008] RACE TO JUDGMENT 109
Moreover, the case had elements that played into stereotypes and standard
scripts about race and class—dynamics that were reinforced by Duke’s mostly
affluent, mostly white student body and faculty located in Durham, a small city
(population approximately 190,000) with a predominantly working-class and
poor population, forty-four percent of which is African American. Some black
leaders and citizens took up and pushed the case in light of its perceived racial
dynamics and town–gown tensions. Faculty and outside groups who saw this
alleged crime as symbolic of real race, class, and gender injustices—which
turned out not to be represented by this particular case—helped propel it
further. In addition, the story played into another cultural schema, one about
elite college sports, athletes behaving badly, and universities doing little to stop
it. Furthermore, Duke is a kind of “celebrity” university, one of a handful that
combines prestige for its academics (ranking eighth in the 2008 U.S. News and
World Report rankings)
70
with highly publicized, championship sports teams—
as personified by “Coach K,” the men’s basketball coach, who does
commercials for American Express and General Motors, among others. In this
sense, the lacrosse saga possessed an allure roughly analogous to that of the
O.J. Simpson murder or the Michael Vick dog-abuse cases. All this may help to
explain why the case obtained so much publicity and also some of the specific
themes that emerged in coverage.
71
A. Sample and Coding for Systematic Quantitative Analysis
Our analysis focuses on stories substantively dealing with this case in The
New York Times, USA Today, the Raleigh News & Observer, and NBC Nightly
News. This focus allows comparison of national newspaper outlets (The New
York Times and USA Today) with a local newspaper (the Raleigh News &
Observer) and comparison of newspaper coverage with network news.
72
We
examined news coverage from March 14, 2006, through April 15, 2007. This
period encompasses the initial accusation and runs until four days after all the
charges were dismissed—through the Sunday following the dismissal of charges,
so as to include the Sunday papers.
73
The sample of substantive coverage dealing
70. U.S. NEWS & WORLD REPORT, AMERICA’S BEST COLLEGES 2008, available at
http://colleges.usnews.rankingsandreviews.com/usnews/edu/college/rankings/brief/t1natudoc_brief.php.
71. Between the time of the event in March and the end of May 2006, CBS Evening News spent 30
minutes on the story, ABC World News 27½ minutes, and NBC Nightly News 21½ minutes. Few if any
events in Durham’s history (aside from Duke basketball) have received as much concentrated,
national-media attention. Yet it is important not to overstate the case. In May 2006, only 16% of the
public said they were following this story very closely, compared with 42% who said they were
following news about the situation in Iraq very closely, 69% who said they were following high gasoline
prices very closely, and 44% who said they were following immigration very closely. P
EW RESEARCH
CENTER FOR THE PEOPLE AND THE PRESS, PUBLIC ATTENTIVENESS TO NEWS STORIES: 1986–2006,
http://people-press.org/nii/bydate.php.
72. Neither cable news nor talk shows were included in the systematic content analysis, though
they will be discussed briefly below. Nor was the local Durham paper included, a paper that has come
under criticism for its coverage. See T
AYLOR & JOHNSON, supra note 1, at 65.
73. This sample was based on a LexisNexis search for all articles that included “Duke” within the
same paragraph as “lacrosse” during this time period.
This initial search generated several stories that
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
110 LAW AND CONTEMPORARY PROBLEMS [Vol. 71:93
with the case includes 86 articles from The New York Times, 40 from USA
Today, 321 from the News & Observer, and 21 stories on NBC Nightly News.
74
Figure 1 shows the amount of substantive coverage over time for this period
for each of these sources. Coverage waxes and wanes with events in the case; it
is always higher in the local paper, as one would expect. The national news paid
considerable attention in the very early period as the case initially unfolded and
indictments were issued (March 2006 through May 2006), returned to the case
when the rape charge was dropped and Nifong recused himself (December 2006
and January 2007), and gave considerable attention when the students were
cleared of all charges (April 2007).
Figure 1: Media Coverage of the Duke Lacrosse Case By Outlet: March 14,
2006, through April 15, 2007
did not substantively deal with the Duke lacrosse case; those with fewer than three paragraphs about
the incident were eliminated. Dropping these search results eliminated stories only peripherally dealing
with the case—such as those about the lacrosse team’s new fall season, their new coach, and the
women’s lacrosse team—while also removing one-sentence news summaries or other articles making
only a passing reference to the case. All letters to the editor and summaries describing an article inside
the paper were also removed from the sample. If an article ran less than three paragraphs, it was
included if the entire article dealt with the lacrosse case.
74. In all, NBC devoted thirty-four minutes of coverage to the case. Its depictions tended to track
major developments—there is no coverage between the indictment of the third defendant in May 2006
and the news that the rape charges had been dropped on December 23, 2006. This is somewhat less
time than the other two mainstream networks devoted to the story: ABC allocated just over forty-one
minutes to coverage of this case over this entire period, and CBS devoted nearly one hour to the case,
including an unusually lengthy ten-minute story on April 11, 2007.
0
5
10
15
20
25
30
Mar 1
9
- Ma
r
2
5
Apr
9
-
Apr 15
Apr 30 - May 6
M
a
y
21 - M
a
y 27
Jun 11 - J
u
n
1
7
Ju
l
2
-
Ju
l
8
Jul 23 - Jul 29
Aug 13 - A ug 19
Sept
3
- Sept 9
Sept
2
4 -
Sept 30
O
c
t
1
5
-
Oct 21
Nov 5
- Nov 11
N
ov
26
-
Dec
2
De
c
17 - De
c
23
J
an 7
-
Ja n 13
J
an
2
8
-
F
eb 3
Feb
18
- Feb
24
M
ar 1
1
- M
ar 17
Apr
1 - A
pr 7
Number of Stories
Ne w York Time s USA Today NBC Nightly News Ra l eigh News & Obse rver
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
Autumn 2008] RACE TO JUDGMENT 111
In characterizing the nature of coverage, the coding strategy was two-fold.
First, it included examining the total number of substantive articles invoking
specific terms that might contribute to favorable or unfavorable evaluations of
the accuser and the accused, in addition to examining how often articles put the
coverage into a racial or class context to understand the larger frames used in
making sense of this event.
Second, the coding strategy included analyzing the degree to which news
stories provided information that would promote favorable or unfavorable
judgments of those associated with the prosecution and those associated with
the defense—that is, the degree to which the information would promote, in a
neutral audience member, a belief either in the defendants’ guilt or in their
innocence. Specifically, each paragraph was coded for the presence or absence
of information that might move a neutral reader toward the belief that the
players were guilty or toward the belief that they were innocent. Every
paragraph was thus coded as containing (1) only information that contributed to
the belief the players were guilty, (2) only information that contributed to the
belief the players were not guilty, (3) information that promoted both guilty and
nonguilty views, or (4) no information that might move a neutral reader in one
direction or the other.
75
This analysis provides a systematic basis for inferences
about whether, when, or where the coverage did tilt heavily against the
presumption of innocence.
B. Results
1. Racial and Class Context
Usual media behavior would lead one to expect simplifying and
sensationalizing—the media rely heavily on stereotypes in constructing their
narratives and journalists themselves filter information through their own
stereotypes. In the Duke lacrosse case, there was evidence that many in the
media used a standard story script that viewed this event as a metaphor for
larger cultural divisions—white over black, the privileged over the working
class.
76
We searched for the terms “race,” “racial,” “racism,” “racist,” “black,”
“white,” and “African American” in our samples of substantive articles,
assuming these words signal that the story implicates race. Table 1 reveals that a
majority of the national news coverage included one or more of these terms—
57% of The New York Times articles, 80% of USA Today articles, and 48% of
75. For purposes of this analysis, a random sample of one-half The New York Times substantive
articles (n=44) and a random sample of one-fifth of the News & Observer substantive articles (n=69)
were coded. Because there was less coverage, every USA Today substantive article and every NBC
Nightly News story was coded. See Appendix for further details on the coding.
76. The setting—prestigious university with a white-majority student body in a town where the
majority of the residents are black or working class—may have contributed to this metaphor even while
the media overstated the case. As community, faculty, and outside groups took up the case, they often
did so from perspectives that advanced this narrative.
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
112 LAW AND CONTEMPORARY PROBLEMS [Vol. 71:93
NBC Nightly News stories. The News & Observer used racial frames somewhat
less: 43%.
This analysis suggests that media coverage did place the case in a racial
context, particularly among national news organizations unfamiliar with the
local context. Framing the case through a racial lens—a choice not somehow
compelled by the laws of nature, but rather by the normal practices of
professional journalism—might yield both benefits and harms. On the one
hand, by racializing the story, journalists convey the role that racial difference
continues to play in the United States. To some degree, that reactions emerged
as they did suggests race was indeed an important force shaping belief and
behavior in the case. On the other hand, this narrative, like many others, places
race in a context where racial hostility and competition are highlighted and
perhaps exaggerated. Such information is likely to undermine racial comity.
77
National news outlets emphasized the race angle more than the Raleigh paper.
Perhaps more-detailed knowledge of the local community enables local
journalists to grasp how racialized stereotypes and simplifications can possess
elements of truth while also misleading readers. Additionally, it was the racial
element that helped motivate national-media and audience attention to this
local criminal case in the first place.
77. See generally ENTMAN & ROJECKI, supra note 18, at 11–12 (“[R]acial comity means something
more than mere tolerance . . . [:] [it] require[s] that members . . . act kindly and empathetically enough
to see beyond skin color to their own shared interests in a more effective and harmonious society.”).
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
Autumn 2008] RACE TO JUDGMENT 113
Table 1: Articles with Racial and Class References
78
The New
York
Times
USA
Today
NBC
Nightly
News
Raleigh
News &
Observer
Articles and stories in which any of
the terms “race,” “racial,” “racist,”
“racism,” “black,” “white,” or
“African American” appear
57%
(n=49)
80%
(n=32)
48%
(n=10)
43%
(n=138)
Articles and stories in which only the
terms “racism,” “racist,” or “racial
hatred” appear
14%
(n=12)
13%
(n=5)
10%
(n=2)
10%
(n=32)
Articles and stories with references
to “racial epithets,” “insults,” “slurs,”
or “racial taunts” by the lacrosse
players
21%
(n=18)
18%
(n=7)
5%
(n=1)
14%
(n=44)
Articles and stories with references in
which the terms “class,” “privilege,”
“well-off,” “well-to-do,” “affluent,”
“suburban,” “poor,” “working class,”
or “poverty” appear
35%
(n=30)
40%
(n=16)
29%
(n=6)
18%
(n=58)
Total stories N=86 N=40 N=21 N=321
There is less evidence that national news media explicitly portrayed the
players as “racist,” though there was certainly attention, particularly early on, to
the claim that the players at the party used racial slurs. Evidence on this point
can be found in the second and third row of Table 1. The terms “racism,”
“racist,” or “racial hatred” appear in relatively few articles (this group is a
subset of the group reflected in the second row). In the context in which they
appear, the terms were very rarely used to explicitly label the players as racist.
79
78. Sources: The New York Times, USA Today, the Raleigh News & Observer, and NBC Nightly
News, from March 14, 2006, through April 15, 2007, based on LexisNexis search with terms “Duke” and
“lacrosse” in the same paragraph, discarding articles that are not substantively about the case. Table
entries report the proportion of all substantive stories that invoke at least one of the terms in the story.
79. These terms were used relatively infrequently in referring to the students themselves. Take for
example the five USA Today articles that invoked these terms. One referenced “racist jocks,” citing
Tom Wolfe’s novel I am Charlotte Simmons. Sal Ruibal, Assault Scandal Highlights Divide for Durham,
Duke, USA
TODAY, Mar. 13, 2006, at 9C. Another, appearing after the first indictments, did reinforce
the claim that racism was a consideration in the event. It reported that students at North Carolina
Central University (NCCU) had “racist graffiti spray-painted on doors and . . . had racist comments
shouted at them by people in passing cars.” Dick Patrick, N.C. Central Applauds Action; Students Hope
Justice Served, USA
TODAY, Apr. 19, 2006, at 10C. It quoted an NCCU professor who was monitoring
the case for the state chapter of the NAACP, and it went on to quote a student who said people need to
understand and discuss these issues of “racism, sexism, [and] power.” Id. A third story offered two
alternative historical references for the event: Tawana Brawley, who made up a story that she was
raped by “white racists,” and the case of the St. John’s lacrosse players, in which an allegation of sexual
abuse turned out to be true, though the defendants were acquitted of criminal charges. DeWayne
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
114 LAW AND CONTEMPORARY PROBLEMS [Vol. 71:93
Rather, this claim was made more implicitly through reference to the use of
racial epithets by the players at the party. Eighteen New York Times articles
included the terms “racial epithet,” “insult,” “slur,” or “taunt”; seven USA
Today articles, forty-four News & Observer articles, and two NBC Nightly News
stories invoked these terms.
80
The bottom row of Table 1 provides evidence on the extent to which the
case was placed into a class frame. Here the search comprised terms that might
signal class as a frame of reference: “class,” “privilege,” “well-off,” “well-to-
do,” “affluent,” “suburban,” “poor,” “working class,” and “poverty.” Although
not as prominent as a racial frame, a significant portion of substantive coverage
included at least one of these class terms: 35% of articles in The New York
Times, 40% in USA Today, 18% in the News & Observer, and 29% on NBC
Nightly News.
Whereas the race and class storyline was prominent, very few stories placed
the incident in a gendered context. Only eight Times stories invoked gender, six
in USA Today, seventeen in the News & Observer, and two on Nightly News.
81
Wickham, Race and Sex Cast Long Shadow Over Duke, USA TODAY, Apr. 18, 2006, at 13A. In the
spring of 2007 these terms appeared in two more articles. A sports column suggested that the three
defendants were “part of a group that included vile, racist teammates,” Jon Saraceno, Duke Lacrosse
Case Has No Winners, USA
TODAY, Feb. 26, 2007, at 5C, and an article quoting editorials and columns
from across the nation criticized the faculty for supporting the branding of the athletes as “racist,” A
Rush to Judgment on Duke Lacrosse Men, USA
TODAY, Apr. 13, 2007, at 11A. On NBC Nightly News,
the first story on the topic opened with, “And on the campus of one of the nation’s top universities,
Duke, accusations of racial hatred and a terrible crime . . . .” NBC Nightly News (NBC television
broadcast Mar. 29, 2006). The second occurred in a story noting that “[b]ecause the accuser [was]
African-American, the issue of racism inflamed the case and heightened tensions . . . .” NBC Nightly
News (NBC television broadcast Apr. 10, 2006). The New York Times used these terms in twelve
articles. In six, the terms were used in ways that undercut the claim of racism. An article from April 12,
2007, that began by reporting that the three students had been declared innocent of all charges also
noted “the accuser’s vivid account of racism and misogynistic taunts.” Duff Wilson & David Barstow,
Duke Prosecutor Throws Out Case Against Players, N.Y.
TIMES, Apr. 12, 2007, at A1. This story also
offered a quote from a student suggesting it was unfair to pigeonhole all students as racist. Id. One
column about Reade Seligmann, one of the accused students, includes a quote from the only black
member of the women’s lacrosse team who, in response to the question whether he ever gave indicators
of being racist said, “Oh my god, absolutely no, a resounding no. The idea is just laughable.” Peter
Applebome, As Duke Accusation Festers, Disbelief Grows, N.Y.
TIMES, July 16, 2006, § 1, at 23. An
article and a column cite the faculty report that notes no evidence that the lacrosse players were racist.
David Brooks, The Duke Witch Hunt, N.Y.
TIMES, May 28, 2006, § 4, at 11; William Yardley, Review by
Duke Faculty Sees Both Bad and Good in Lacrosse Team, N.Y.
TIMES, May 2, 2006, at A22. In an early
article, a lawyer for the players said that they had been branded racist (the implicit message being that
they were not). Joe Drape, Lawyers for Lacrosse Players Dispute Accusations, N.Y.
TIMES, Mar. 31,
2006, at D5. And finally, a column by Nicholas Kristof argued that racism runs through history but that
this should prompt deep reflection not just on racism but on prejudice of any kind, including prejudice
against athletes. Nicholas Kristof, Op-Ed, Jocks and Prejudice, N.Y.
TIMES, June 11, 2006, § 4 at 13.
Note, however, that athletes (particularly white ones) have never been subjected to systematic (or
centuries-long, government-sanctioned) discrimination or violence rooted in a negatively stereotyping
ideology.
80. Included in this are a few references to the allegation that Finnerty used a slur in an assault on
a man in Washington, D.C. Over half of all references to the use of racial insults occurred in March and
April of 2006.
81. Here, the search terms were “sexist,” “sexism,” “sex,” or “misogynist.” All stories that were
essentially referring to the allegations of sexual assault were dropped. In the Times, when a gendered
context was invoked it was often alongside a racial context—for example, “another painful chapter in
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
Autumn 2008] RACE TO JUDGMENT 115
Presumably, media motivated by “liberal bias” or desire to promote a
“politically correct” agenda would have made more of the patriarchal ritual
embodied in the hiring of a stripper by young men, irrespective of rape charges.
2. Terms Used in Describing the Accuser and the Accused
How the press characterized the accuser was also explored
82
—specifically,
how often she was described as a mother and how often she was described as a
student (assuming these labels facilitate more-sympathetic impressions). The
words “mother,” “student,” “single mother,” or “North Carolina Central” were
searched for; references not associated with the accuser were dropped.
“Stripper,” “exotic dancer,” and “dancer” were also searched for; references
not linked to the accuser were likewise dropped.
The results can be found in Table 2. A small but significant number of
articles did describe the accuser as a mother (or single mother) and student.
This positive depiction is likely offset by the fact that she was much more likely
to have been described, given the facts of the case, as a stripper or dancer. Here,
television news proves to be an exception—almost as likely to have described
her as a student as to have described her as a stripper. Perhaps the difference
arose because network television news paid little attention to the case between
the initial phase (when “student” was more common) and the end (when
“stripper” dominated).
References to the accuser as a mother and student are indeed concentrated
in early coverage, which tended to treat her more sympathetically. In fact, the
number of references that described her as a student is similar to the number of
references that noted she was a stripper, up to the point at which the initial
DNA evidence was released. Subsequently, descriptions of her as a student
dropped off more rapidly than descriptions of her as a stripper. Over time, as
evidence mounted suggesting the players were innocent, she was described
mainly as a stripper.
This analysis of how she was described or identified does not suggest she
was treated especially gently by the media. Of course, it is difficult to know how
audiences interpreted these references to her occupation. These labels have the
the tangled American opera of race, sex, and privilege”; “the case has touched historically tender
nerves of race, sex, and class”; and “pointing to underlying issues of race, class, sex, and privilege.”
82. The news sources analyzed here were not more likely to call her a victim than to call her the
accuser. To the degree that use of the term victim implies that something did happen, potentially
undermining the presumption of innocence, the news did not reinforce this perception by using victim
more than accuser. A simple word search for “accuser,” “victim,” and “alleged victim” shows the
papers significantly more likely to describe her as the accuser. However, when she was referenced as a
victim, it was not as an alleged victim. The New York Times used the term “accuser” 151 times
compared with 38 times for “victim” (only once was the term “alleged victim” used). In the News &
Observer the term “accuser” appeared 627 times in coverage, “victim” 150 times, and “alleged victim”
24 times. USA Today and NBC Nightly News were relatively more likely to use “victim.” In USA
Today, “accuser” appeared 49 times, “victim” 18 times, and “alleged victim” 3 times; on NBC Nightly
News these figures were 17 times, 9 times, and 6 times respectively.
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
116 LAW AND CONTEMPORARY PROBLEMS [Vol. 71:93
potential to undermine audience sympathy,
83
particularly in light of the
consistent references to her having been heavily intoxicated on that night. On
the other hand, such labels also reinforce the negative image of the students by
reminding audiences that they hired strippers for an underage drinking party.
Table 2: Articles that Include Specific Terms in Referencing the Accuser
84
The New York
Times
USA Today
NBC Nightly
News
Raleigh News
& Observer
“Mother” 15% (n=13) 10% (n=4) 10% (n=2) 3% (n=10)
“Single Mother” 9% (n=8) 5% (n=2) — 3% (n=9)
“Student” 41% (n=35) 40% (n=16) 23% (n=5) 11% (n=36)
“Stripper” or
“Dancer”
56% (n=48) 85% (n=34) 29% (n=6) 55% (n=176)
Total stories N=86 N=40 N=21 N=321
The analysis also focused on how often “hooligan,” a term Nifong applied to
the players, appeared in the coverage, finding that it was presented mostly in a
critical context. Five of nine articles that used this term in The New York Times
made clear the label was unjustified, or criticized Nifong for having used it.
Similarly, five of six references in USA Today and two on NBC criticized
Nifong’s terminology. In the News & Observer, nineteen of twenty-five stories
containing “hooligan” undermined its applicability. So, in thirty-one of forty-
two articles (seventy-four percent), the coverage told audiences that the players
were not hooligans or, what amounts to the same thing, that Nifong should not
have used the word.
The damaging nature of certain keywords such as “single mother” and
“stripper” must also be considered in any calculation of the balance between
coverage of the defendants and coverage of the accuser. And this is a good
example of the complexity of trying to decipher the polysemic nature of the
coverage and its effects. It is unclear, for example, whether denunciations of the
lacrosse players as hooligans caused any more damage to their reputations—no
matter how subjectively painful these characterizations may have been to the
83. She was a black, single mother working as a stripper while raising young children (a portrait
consistent with negative stereotypes about black women). And to the extent that some of the articles
detail her work as a stripper, they tend to paint an unflattering portrait.
84. Sources: The New York Times, USA Today, NBC Nightly News, and the Raleigh News &
Observer, from March 14, 2006, through April 15, 2007. This table presents the percentage of stories in
which the accuser was described in each of these ways. Coding for “single mother” is done separately
from coding for “mother.”
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
Autumn 2008] RACE TO JUDGMENT 117
players and their families—than being labeled as an unmarried mother and
stripper damaged the reputation and credibility of the accuser.
Although the “hooligan” label was largely dismissed, there is some support
for the claim that early coverage contained information that reinforced
impressions of Duke lacrosse players as spoiled, privileged, and loutish. There
are references to prior misdemeanor charges, quotes from neighbors and
professors, and discussions of the team’s collective reputation that can be
characterized as undermining the individual defendants’ reputations. For
example, The New York Times noted how fifteen members of the team had
committed previous minor infractions, like public intoxication, in a story
headlined, “A Team’s Troubles Shock Few at Duke.”
85
Although our analysis
does not comprehensively evaluate this dimension of the coverage, some
suggestive evidence comes from a search for the terms “record,” “criminal,”
“violat[-e, -ing, -ion, -or, -ed],” or “assault.” This search sought any mention of
a criminal record or of violating university codes for either the accuser or the
accused, dropping any irrelevant references. In The New York Times, two
articles mentioned the accuser’s record and ten mentioned the defendants’
records. In USA Today, these figures were one and six; in the News & Observer,
twenty-three stories referred to the defendants’ records and five to the accuser’s
prior record. NBC Nightly News stories contained only a single reference to a
prior charge against one of the defendants.
86
Viewing this particular dimension
as a frame contest, the score would be eight to forty against the defendants. On
the other side, however, the news also offered many positive character
witnesses testifying to the decency of these young men and relatively few
testimonials to the character of the accuser.
3. The Use of DNA Evidence
Because DNA evidence played such a prominent role in the case—not least
because Nifong originally asserted that it was crucial and then backtracked
when tests failed to implicate members of the team—we examine how the
national media outlets presented this evidence, as detailed in Table 3. Here, we
searched for all articles that included “DNA” and then analyzed whether the
references were exculpatory or incriminating. In The New York Times, eighty-
three references to exculpatory DNA evidence appear in forty-three stories,
and seven references to incriminating DNA evidence appear in six stories. The
latter were mainly references to a second test on the accuser’s fingernails that
had a match with the third player indicted. USA Today mentioned exculpatory
DNA evidence twenty-seven times in eighteen articles and NBC Nightly News,
eleven times in seven stories. Totaling the numbers reveals 121 mentions of
85. Warren St. John & Joe Drape, A Team’s Troubles Shock Few at Duke, N.Y. TIMES, Apr. 1,
2006, at D1.
86. Articles describing past violations of the defendants tended to mention Finnerty’s assault
charge in Washington, D.C., that Evans went to court for a noise violation, and a general record of
misconduct on campus among lacrosse players.
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
118 LAW AND CONTEMPORARY PROBLEMS [Vol. 71:93
exculpatory DNA versus eleven mentions of incriminating DNA, a strong
imbalance—reflecting the facts—favoring the defendants.
87
Looking only within
articles that mentioned the term “DNA” further reveals thirty-three mentions
of Nifong’s continued belief in guilt or of the medical exam having been
consistent with rape, which might offset claims that the DNA evidence
exonerates the defendants. By these measures, attention to the DNA tests’
exculpatory results thoroughly prevails (by 121 to 44) over attention to
indications of guilt.
87. Looking at when these references appear reveals—as would be expected—that most mentions
came after the initial DNA results were released on April 10, 2006. More interestingly, mentions of
exculpatory DNA evidence are fairly evenly divided between coverage in the spring and summer and
coverage in the fall and winter, once it was revealed that Nifong withheld information about additional
exculpatory DNA evidence (presence of other DNA not associated with any lacrosse players).
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
Autumn 2008] RACE TO JUDGMENT 119
Table 3: References to DNA Evidence in National News Outlets
88
DNA
Exculpatory
DNA
Incriminating
DNA
Inconclusive
Assertion that
Medical Exam
is Consistent
with Claims of
Accuser or
Assertion
Offsetting
Claim from
Nifong, or both,
in an Article
Mentioning
DNA Evidence
The New York
Times
Total Mentions
Articles with at
least one reference
83
43
7
6
2
2
25
15
USA Today
Total Mentions
Articles with at
least one reference
27
18
—
—
2
1
5
4
NBC Nightly News
Total Mentions
Articles with at
least one reference
11
7
4
2
2
1
3
2
4. Messages Reinforcing Innocence and Guilt
In this analysis, the most direct test of the extent to which media coverage
favored one side or the other came from a systematic examination of assertions
facilitating claims of guilt or innocence. The results of this analysis are
presented in Table 4. The table reports the proportion of all paragraphs in a
given outlet that can be characterized as leading a neutral reader toward
believing the accused were guilty or innocent. The first row under each source
reports the results for coverage over the full period examined (March 14, 2006,
through April 15, 2007). In each case, when one accounts for all coverage,
88. Sources: The New York Times, USA Today, and NBC Nightly News March 14, 2006, through
April 15, 2007, based on a LexisNexis search for stories with “Duke w/p lacrosse” and “DNA.” The
first row reports total mentions, the second row reports articles with at least one mention, and the final
column reports when information that might offset the exculpatory DNA evidence appears in stories
that mention DNA evidence.
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
120 LAW AND CONTEMPORARY PROBLEMS [Vol. 71:93
paragraphs that reinforced the claims of innocence outnumber those that
reinforced guilt. All the outlets provided substantially more paragraphs purely
promoting conclusions of innocence than encouraging either guilt conclusions
or continued uncertainty (by including both guilt and nonguilt information). In
The New York Times, 21% of all paragraphs coded included only information
or claims that would lead a neutral reader to view the defendants as innocent,
11% promoted guilt, and 4%, uncertainty. In USA Today, the equivalent
figures were 16%, 10%, and 5%, and on NBC Nightly News, 18%, 9%, and
11%. The News & Observer offered a lower proportion of direct innocence or
guilt assertions, with 12% of all coded paragraphs including only information
suggesting innocence, 8% guilt, and 2% uncertainty.
89
Interestingly, The New York Times, which received substantial criticism for
its purportedly pro-Nifong and anti-accused coverage, actually more
disproportionately favored “not guilty” inferences than the other sources,
considering the coverage in its entirety. And averaging across the four media
outlets, 17% of paragraphs were purely pro-innocence and 9.5% purely favored
guilt, nearly a two-to-one ratio favoring the defendants. Certainly these figures
do not portray a media homogeneously attacking the lacrosse players.
Because much criticism focused on early coverage, the analysis is divided
into two periods: on or before April 10, 2006—the day that initial results
showing no DNA match to the defendants were released—and after that day.
90
When the results from the earlier phase are compared to the results from the
later phase, the importance of this distinction emerges.
91
In the initial phase, the
proportion of paragraphs containing information that would be read by a
neutral reader as reinforcing claims of guilt substantially outnumbered those
that reinforced claims of innocence in each outlet. Coverage of this nature
would have the potential to undermine the presumption of innocence,
consistent with the claims of critics. Even in this early phase, however, some
paragraphs supported innocence.
92
89. Front page stories are also no more likely to emphasize guilt. The three newspapers were
combined to compare front page stories (B section page 1 stories from the News & Observer were
included here since it was a local story) with inside stories. Among front page stories, 8% of paragraphs
contained information that only reinforced guilt, 16% of paragraphs contained information that only
reinforced innocence, and 3% contained both. For stories that did not originate on the front page, the
proportions were 10%, 15%, and 3%, respectively. Thus, in the most prominent stories, innocence was
the dominant theme—another finding incompatible with a hypothesis of liberal bias driving the
coverage.
90. In the case of television, the results are presented in the April 10th broadcast, so we place April
10th in the second phase of coverage.
91. See Table 4, Rows 2, 3, 7, 8, 12, 13, 17 & 18 for results divided by time period.
92. See, e. g., Joe Drape, Lawyers for Lacrosse Players Dispute Accusations, N.Y.
TIMES, Mar. 31,
2006, at D5.
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
Autumn 2008] RACE TO JUDGMENT 121
Table 4: Media Coverage Reinforcing Guilt or Innocence, By News Outlet
93
Guilty
Only
Not
Guilty
Only
Both Guilty
and Not
Guilty
Total
Paragraphs
USA Today (n=40 articles)
Total Coverage
On or Before April 10, 2006
After April 10, 2006
News Coverage
Editorials, Columns, and
Commentary
10%
11%
10%
10%
11%
16%
4%
19%
14%
21%
5%
2%
6%
4%
7%
634
151
483
512
122
The New York Times (n=44
articles)
Total Coverage
On or Before April 10, 2006
After April 10, 2006
News Coverage
Editorials, Columns, and
Commentary
11%
9%
13%
13%
3%
21%
2%
29%
23%
6%
4%
1%
5%
4%
—
936
269
667
815
121
Raleigh News & Observer (n=69
articles)
Total Coverage
On or Before April 10, 2006
After April 10, 2006
News Coverage
Editorials, Columns, and
Commentary
8%
15%
7%
9%
2%
12%
5%
13%
11%
16%
2%
1%
2%
2%
2%
1388
127
1261
1193
195
NBC Nightly News (n=21 stories)
Total Coverage
Before April 10, 2006
On or After April 10, 2006
9%
18%
7%
18%
3%
22%
11%
—
13%
198
34
164
93. Sources: The New York Times, USA Today, NBC Nightly News, and the Raleigh News &
Observer, March 14, 2006, through April 15, 2007. In the case of The New York Times and the News &
Observer, the analysis is based on a subsample (randomly selected) of all substantive coverage. Table
entries show the proportion of all paragraphs devoted to the Duke lacrosse case that fall into a specific
category. Paragraphs were coded based on whether they included information, evidence, or claims that
reinforced guilt only, reinforced innocence only, reinforced both, or had no information that could
move a neutral participant toward one side or the other.
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
122 LAW AND CONTEMPORARY PROBLEMS [Vol. 71:93
Given how the proportion of paragraphs reinforcing guilt or innocence
changed over time, it is apparent that the weakness of the prosecution’s case
was not fully realized until well into this second phase. What accounts for this
continuing attention to information that suggests guilt after the initial DNA
evidence came back—evidence that Nifong said would be crucial and that
defense lawyers said would prove their clients’ innocence? Recall that Nifong
indicted the players after April 10th, so the indictments themselves seem to be
evidence reinforcing guilt early in the second phase. Because this DNA
evidence did not mark an unraveling of the case for the prosecutor, journalistic
routines and norms ensured that it did not mark an immediate unraveling of the
case for journalists. Rather, reporters and editors apparently assumed that
Nifong had other good evidence for proceeding with the indictments, and this
colored their reporting.
94
Here one sees how dependence on elite sources helped
maintain attention to information suggesting guilt, even after evidence severely
undermining the prosecutor’s case became public.
As suggested earlier, pretrial publicity is likely to be at least somewhat
slanted in favor of the prosecution, and the early coverage in this case was no
exception. To the extent that criminal defendants ever enjoy a chance to have
their sides presented in equal measure to the prosecution’s, the Duke lacrosse
players occupied a more promising position than most because of their class
status and the presence of some compelling evidence for innocence (and one
sees this realized in the second phase). Early on, however, this advantage was
undermined by Nifong’s vivid assertions of guilt, by claims suggesting the team
was not cooperating, and by news of many team members’ previous uncouth,
though not seriously illegal behavior (much of it well-documented), which
seemed to confirm suspicions and stereotypes. Revelations that one member of
the team had an assault charge in Washington, D.C., and that another member’s
e-mail, sent right after the party in question, said that he wanted to brutalize
and kill strippers, only added to the negative image. First impressions establish
stereotypes that are difficult to dislodge. The early events helped establish these
unfavorable perceptions among audiences as well as journalists, perceptions
that may have stuck even after the initial DNA evidence became available.
95
Eventually, coverage did change from its pro-prosecution slant. Overall, the
second phase is marked by greater attention to the claims of innocence.
96
This is
94. See Rachel Smolkin, Justice Delayed, 29 AM. JOURNALISM REV., Aug.–Sept. 2007, at 18, 25–
27.
95. On the cognitive psychology demonstrating the stickiness of stereotypes, and in particular the
tendency of people to disregard later, stereotype-disconfirming information while selectively noticing
any new, stereotype-confirming data, see Myron Rothbart & Oliver Johnn, Intergroup Relations and
Stereotype Change, in P
REJUDICE, POLITICS, AND THE AMERICAN DILEMMA (Paul M. Sniderman et
al. eds., 1993).
96. In some instances, the nature of the coding scheme used herein probably led to extra tallies on
one side. This occurred toward the end of the sample period, when stories recounted the entire case. In
these instances, reiterations of the accuser’s claims are counted as information reinforcing guilt if the
coded paragraph (or the paragraph immediately preceding it) did not refute the rape charge. Thus, the
paragraphs coded as invoking guilt in the spring 2007 coverage often appeared in the context of stories
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
Autumn 2008] RACE TO JUDGMENT 123
particularly extreme in the period from December 22, 2006, when the rape
charges were dropped, through April 11, 2007, when all charges were dropped
(encompassing the period when Nifong asked to be removed from the case and
the North Carolina Attorney General took over). As Figure 1 makes clear,
there was considerable media attention during this time. In this period, the
results are as follows: in the News & Observer, 19% of paragraphs emphasized
only innocence and 5% emphasized only guilt; in The New York Times, the
figures are 52% innocence and 6% guilt; in USA Today, 40% innocence and
6% guilt; and on NBC Nightly News, 38% innocence and 3% guilt.
97
5. Editorials Versus News
Observations of media coverage of the Duke lacrosse case suggest that far
and away the worst offenders in framing from a one-sided prosecution
perspective were certain opinion columnists and cable television personalities.
These people are not reporters bound (at least nominally) by the strictures of
“objectivity” or balance; nor do they necessarily even face the requirement of
confirming facts or of doing any reporting of their own. One example is a
“Sports of the Times” column by Selena Roberts in which she wrote about a
“night that threatens to belie [the team members’] social standing as human
beings” and in which she clearly implied the students’ guilt.
98
Another example
would be a Sunday op-ed piece, Blue Devils Made Them Do It, which presented
a highly unfavorable portrait of the Duke lacrosse team and again did little to
apply the innocent-until-proven-guilty standard.
99
Here, analysis of the proportion of paragraphs that reinforced guilt or
innocence suggests that editorials were much more likely to be one-sided—that
is, to include only paragraphs reinforcing guilt or only paragraphs supporting
innocence, with no paragraphs at all representing both sides. All editorials and
columns in The New York Times were one-sided in this way, purely supporting
guilt or innocence, as were seven of eleven editorials and columns in the News
& Observer. This said, however, editorials are designed to assert opinion.
Perhaps most egregiously imbalanced were the cable talk shows, such as The
Nancy Grace Show on the CNN Headline News network.
100
Many cable shows
that began with the fact that the defendants had been cleared. The coding protocol does not fully
capture how that initial background might influence interpretation of all subsequent paragraphs; in
other words, the data somewhat overstate the degree to which the defendants’ guilt continued to be
asserted during the second phase.
97. The heavy emphasis on innocence near the very end of coverage in the second phase does not
fully account for the finding of more attention to innocence in the second phase. For example, in The
New York Times during the period from April 10, 2006, through December 22, 2006 (before the
indictments were dropped), 15% of paragraphs emphasized only guilt and 22% emphasized only
innocence.
98. See Selena Roberts, When Peer Pressure, Not a Conscience, is Your Guide, N.Y. T
IMES, Mar.
31, 2006, at D1.
99. See Allan Gurgangus, Op-Ed, Blue Devils Made Them Do It, N.Y. T
IMES, Apr. 9, 2006, § 4, at
13.
100. See Smolkin, supra note 94, at 21, 25; Y
AEGER & PRESSLER, supra note 2, at 149–54; TAYLOR
& J
OHNSON, supra note 1, at 118–28. It should be noted that Grace does not generally take overtly
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
124 LAW AND CONTEMPORARY PROBLEMS [Vol. 71:93
seek out guests who will take extreme positions and offer controversial claims
because that is part of the “entertainment” of the show. This coverage becomes
problematic if citizens see it as news, or as some form of journalism. As KC
Johnson notes, “They shouldn’t be considered journalism in any respect at all,
but it’s alarming because most of the public probably does consider it
journalism, or journalism in some way.”
101
Of course, to the degree that commentary and editorials influence the
construction of news narratives, they may be consequential. This may be a
dynamic at play in early coverage. Editorials and columns appearing before the
release of the initial DNA evidence, like the two examples given above, tended
to support a pro-prosecution framing of the event, although editorials
supporting the claims of innocence also appeared. Analyzing the proportion of
paragraphs that reinforced guilt or innocence separately for news versus
columns or editorials reveals that over the entire period editorials were more
likely to contain paragraphs emphasizing only innocence than they were to
contain paragraphs emphasizing only guilt (Table 4). Overall, editorial pages
averaged 5.3% of paragraphs arguing for guilt only and 14.3% supporting
innocence only, a nearly three-to-one ratio favoring the defendants. In other
words, once strong (DNA) evidence of innocence was available, the newspapers
in the sample gave far more editorial attention to innocence than they ever gave
to guilt. This finding offers another indicator that neither a “politically correct”
nor a liberal editorial agenda was driving the coverage.
V
D
ISCUSSION
This analysis of the Duke lacrosse case coverage in three national news
outlets and one local newspaper reveals evidence consistent with some of the
major concerns raised in critiques of the media’s treatment of the case. The
results provide some evidence that journalists relied on stereotypes and scripts
that may have oversimplified and misrepresented what the case was about and
caricatured and misrepresented the individuals involved. Coverage in the early
phase was also less balanced and exhibited a slant toward the prosecution. And
yet, the analysis here does not provide support for those who would
characterize coverage as “politically correct sensationalism,”
102
or as failing to
include claims of and evidence for innocence. Coverage was not completely
one-sided and unfavorable to the students. This examination suggests that the
liberal positions on her show, so the slanted coverage should not be interpreted as an indicator of
liberal bias. Instead, as noted by Dowler, “The Nancy Grace Show selectively targets specific
kidnappings, sex crimes, or murders, with a particular focus on retribution and punishment.” Dowler et
al., Constructing Crime: Media, Crime, and Popular Culture, 48 C
AN. J. CRIMINOLOGY & CRIM. JUST.
837, 838 (2006).
101. Smolkin, supra note 94, at 25. See generally Robert Entman, The Nature and Sources of News,
in T
HE INSTITUTIONS OF DEMOCRACY: THE PRESS (Kathleen Hall Jamieson & Geneva Overholser
eds., 2005).
102. T
AYLOR & JOHNSON, supra note 1, at 118, use this as the title of a chapter on the press.
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
Autumn 2008] RACE TO JUDGMENT 125
coverage, taken as a whole, does not demonstrate particularly clear-cut
favoritism toward either side. Plenty of unfavorable information was conveyed
about the accuser and the accused.
103
Ironically, just as the earliest coverage may
have biased some people’s perceptions of the case toward guilt, so too the news
and commentary during the first phase may have distorted perceptions of the
media’s behavior. Any conclusions about the journalism here must avoid
stereotypes that oversimplify the media’s constructions of the Duke lacrosse
case.
One of the most important lessons to be drawn is the way normal news
routines—including overreliance on public officials (specifically the prosecution
and police), overuse of familiar stereotypes and standard scripts, and pack
journalism—can combine with a determined government official to facilitate
pro-prosecution slant in news of alleged crimes. The Duke lacrosse case
provides a particularly vivid and compelling example of a more general
dilemma for the administration of justice. Research suggests that pretrial
publicity may have important effects on juries and thus on the ability of
defendants to get a fair trial.
104
The findings are particularly troubling when one
considers that the Duke lacrosse players do not typify crime-news defendants.
Poorer racial minorities—and whites—can rarely afford the sort of legal
representation that helped the Duke students generally obtain more-positive
103. See Susan Hanley Kosse, Race, Riches & Reporters—Do Race and Class Impact Media Rape
Narratives? An Analysis of the Duke Lacrosse Case, 31 S. I
LL. U. L.J. 243, 261, 267 (2007) (concluding,
on the basis of a content analysis of fourteen magazine articles, that the media painted an unflattering
picture of both the accuser and the accused).
104. Much research suggests that pretrial publicity has the potential to produce an antidefendant
bias. See, e.g., John W. Wright II & Susan Dente Ross, Trial by Media?: Media Reliance, Knowledge of
Crime and Perception of Criminal Defendants, 2 C
OMM. L. & POL’Y. 397 (1997) (finding that exposure
to crime coverage is associated with perceptions of a guilty defendant); Gary Moran & Brian L. Cutler,
The Prejudicial Impact of Pretrial Publicity, 21 J. A
PPLIED SOC. PSYCHOL. 345, 345 (1991) (finding that
exposure to pretrial publicity is associated with perceived culpability but not with willingness to admit
one might be biased by the publicity); Stanley Sue et al., Biasing Effects of Pretrial Publicity on Judicial
Decisions, 2 J. C
RIM. JUST. 163 (1974) (claiming that pretrial publicity results in antidefendant bias by
affecting the evaluation of evidence presented at trial); Amy L. Otto et al., The Biasing Impact of
Pretrial Publicity on Juror Judgments, 18 L
AW & HUM. BEHAV. 453 (1994) (suggesting that pretrial
publicity influences initial judgments about guilt); see also sources cited supra note 11. Christine Ruva
et al. suggest a cognitive mechanism by which this might work, arguing that jurors selectively attend to
trial information that is consistent with their preconceived notions based on the pretrial publicity.
Christine Ruva et al., Effects of Pre-Trial Publicity and Jury Deliberation on Juror Bias and Source
Memory Errors, 21 A
PPLIED COGNITIVE PSYCHOL. 45, 61 (2007); see also Lorraine Hope et al.,
Understanding Pretrial Publicity: Predecisional Distortion of Evidence by Mock Jurors, 10 J.
E
XPERIMENTAL PSYCHOL.: APPLIED 111 (2004) (“The effect of prejudicial [pretrial publicity] on
verdict outcomes was mediated by predecisional distortion [juror bias] in the evaluation of
testimony.”); Margaret Bull Kovera, The Effects of General Pretrial Publicity on Juror Decisions: An
Examination of Moderators and Mediating Mechanisms, 26 L
AW & HUM. BEHAV. 43 (2002) (finding
that media slant affected participant standards for determining guilt). Some work suggests that voir dire
and jury instructions do not necessarily overcome these effects. See, e.g., Hedy Red Dexter et al., A Test
of Voir Dire as a Remedy for the Prejudicial Effects of Pretrial Publicity, 22 J. A
PPLIED SOC. PSYCHOL.
819 (1992). See generally Solomon M. Fulero, Afterword: The Past, Present and Future of Applied
Pretrial Publicity Research, 26 L
AW & HUM. BEHAV. 127 (2002) (summarizing the prejudicial pretrial-
publicity literature); J
ON BRUSCHKE & WILLIAM E. LOGES, FREE PRESS VS. FAIR TRIALS:
E
XAMINING PUBLICITY’S ROLE IN TRIAL OUTCOMES (2004) (reviewing prejudicial pretrial-publicity
literature, including many studies finding no effects).
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
126 LAW AND CONTEMPORARY PROBLEMS [Vol. 71:93
coverage after the initial phase. Indeed, for most poorer defendants, there may
be no second phase to balance initial stories almost always slanted against them.
When the defendants are blacks or Latinos, they further suffer from the effects
of pre-existing schemas that arise from prior, racialized coverage of news as
described above.
In these senses, most pretrial publicity is predictably prejudicial, and media
behavior in the Duke lacrosse case was not unusual but typical. Here, the
absence of a second side consisting of legitimate sources promoting an
alternative narrative was compounded by the early public stance of the team, an
almost classic example of poor public relations. Lesson number one for
organizations facing a crisis of confidence (for example, Johnson & Johnson’s
poisoned Tylenol scare
105
) is to create an appearance of openness to criticism
and honest desire to let the truth prevail.
106
Although criminal defendants
obviously differ from corporations or government agencies charged with
malfeasance or error, defense lawyers in recent years have widely recognized
that in high-profile cases they might need to compete in the media to counter
the prosecution bias.
107
To yield balanced coverage, media framing contests
require that both sides have approximately equal skill and resources. The Duke
lacrosse case featured a publicity-hungry D.A. (not exactly unusual) who was
also flagrantly violating legal standards for prosecutors (less typical). On the
other side, it featured a defense-communication strategy initially constrained by
the potential future trial and the involvement of several different potential
defendants and their lawyers. For a few perhaps crucial days, the prosecution
and police dominated media coverage.
108
Furthermore, some journalists suggest that the early silence of team
members, who on advice of counsel refused to speak publicly, inhibited their
105. See generally Jerry Knight, Tylenol’s Maker Shows How to Respond to Crisis, WASH. POST,
Oct. 11, 1982, at Wash. Bus. 1 (discussing Johnson & Johnson’s response); Richard S. Levick, Brand
Protection: Defining Priorities During Crisis and Litigation, http://www.bernsteincrisismanagement.
com/nl/crisismgr040215.html (last visited Feb. 27, 2008) (discussing Johnson & Johnson’s response as
“the ultimate brand preservation”).
106. See, e.g., R
OBIN COHN, THE PR CRISIS BIBLE: HOW TO TAKE CHARGE OF THE MEDIA WHEN
ALL HELL BREAKS LOOSE 182 (2000) (“When a company doesn’t talk to the press, it is giving up
control of the story.”); see also L
YNN BREWER ET AL., MANAGING RISKS FOR CORPORATE
INTEGRITY: HOW TO SURVIVE AN ETHICAL MISCONDUCT DISASTER 149–86 (2006) (discussing crisis-
communication best practices).
107. It would depend on the details of the case whether the defense in high-profile cases would want
to compete in the public arena. See generally John Watson, Litigation Public Relations: The Lawyers’
Duty to Balance News Coverage of their Clients, 7 C
OMM. L. & POL’Y. 77 (2002) (discussing the effect
of duties owed to clients in litigation public-relations strategies); Robert Shapiro, Secrets of a Celebrity
Lawyer: How O.J.’s Chief Strategist Works the Press, 33 C
OLUM. JOURNALISM REV., Sept.–Oct. 1994,
at 25 (discussing litigation public-relations strategies); Jonathan M. Moses, Legal Spin Control: Ethics
and Advocacy in the Court of Public Opinion, 95 C
OLUM. L. REV. 1811 (1995) (discussing the ethical
implications of litigation public relations).
108. The first statements by the Durham police came on March 24, 2006, asserting strong physical
evidence of rape, T
AYLOR & JOHNSON, supra note 1, at 63, and Nifong began making public
statements on March 27, 2006, id. at 85. The first public denial of the rape charge by individuals
associated with the team was a statement released by Duke on March 28, 2006. Id. at 93.
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
Autumn 2008] RACE TO JUDGMENT 127
coverage.
109
To the extent the defense could have contested the early framing,
they undermined their credibility by depriving journalists of the simple raw
material (for example, quotes and interview photos) necessary to balanced
depictions.
110
Early on, Nifong claimed that the players were not cooperating,
and this theme was picked up by reporters and columnists. This, along with
visual images of the team and the paucity of public statements, helped to feed
the impression that “they have something to hide” and “they think they are
above the law.” The team’s so-called “wall of silence,”
111
along with widely
reproduced visual images of team members dressed similarly and appearing
grim faced, potentially left damaging impressions among journalists, and among
audience members predisposed to be sympathetic toward the accuser.
112
Although the team captains initially cooperated by giving statements to the
authorities and volunteering to take polygraph tests, these facts did not fit
conveniently into the initial frame and therefore did not receive narrative
prominence. On the other hand, media research offers little reason to expect
news reports in most arenas to calibrate the magnitude and resonance of
attention given to aspects of the story according to a self-conscious assessment
of those aspects’ substantive or logical importance.
113
In this instance, given
journalistic values, it is not surprising that a symbolic “wall of silence” trumped
the detail that some team members were cooperating with the prosecutor’s
requests.
In some cases, the absence of a second side to the story, due to the lack of a
second set of credible sources promoting a competing frame, arises because
such an argument would be culturally unthinkable or illegitimate. Thus, the
Michael Vick dog-fighting case produced overwhelmingly one-sided coverage
because no credible sources were systematically promoting the arguments that
torturing and killing dogs is an acceptable hobby. Unfortunately, some of the
109. See Smolkin, supra note 94, at 21.
110. See Entman, Framing Bias, supra note 3 (showing through a news-slant index formula the
necessity of two sides competing equally skillfully to yield balanced news).
111. This term was often invoked by editorial writers and columnists. As an indicator of the extent
to which this term entered the lexicon of the case, a simple search of the LexisNexis database for March
14, 2006, through July 31, 2007, finds thirty-three uses of “wall of silence” in U.S. newspapers and
nineteen on ABC, CBS, CNN, or FOX News programs. Many of these, however, came in contexts
denying such a wall existed.
112. On the importance of visual displays of such nonverbal communication behaviors as body
language and facial expressions, see generally Eric P. Bucy & Maria Grabe, Image Bite News: An
Underappreciated Source of Political Information (Aug. 30–Sept. 3, 2006) (paper presented at the 2006
Annual Meeting of the American Political Science Association, Political Communication Division, on
file with Law and Contemporary Problems).
113. See, e.g., B
ENNETT, NEWS: THE POLITICS OF ILLUSION (2007) (showing the vulnerability of
journalists to manipulation by “spin doctors,” who have no interest in logic or substance, but seek
rather to promote news serving their political goals); Robert M. Entman, Mass Media and Policy
Innovation: Opportunities and Constraints for Public Management, in I
NNOVATION IN AMERICAN
GOVERNMENT: CHALLENGES, OPPORTUNITIES, AND DILEMMAS 202 (1997) (showing news-
production values and routines usually trump any interest in calibrating the amount and focus of
coverage to the substantive importance of policy problems and solutions);
Entman, Framing Bias, supra
note 3 (showing that facts and journalists’ assessments of them are less important to the slant of news
than contending parties trying to spin the story).
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
128 LAW AND CONTEMPORARY PROBLEMS [Vol. 71:93
same dynamics can be seen in coverage of military interventions in which fear
of seeming unpatriotic or “soft on defense” discourages credible opponents
from organizing together to promote an alternative narrative suggesting
diplomacy as a more efficient solution to the conflict.
114
In one sense, the
defendants here suffered from a similar cultural disadvantage in that the facts of
the case, as promoted vigorously by the prosecution, seemed indeed culturally
indefensible. However, as eventually became clear, there was another culturally
acceptable perspective—one emphasizing prosecutorial misconduct.
The media might have been more sympathetic to the defense’s questioning
of Nifong’s credibility, had that been more of a focus of the defense’s initial
media strategy; but the defense did not skillfully promote it during the crucial,
early days when the “sticky,” initial, event-specific schema for the incident was
forming.
115
Instead, the defense appears to have spent more time attacking the
accuser’s story and credibility. One reason this tack might have been less
effective is that, had the story turned out to be true, journalists would have
found themselves on the side of attacking a rape victim.
116
The apparent story of privileged whites victimizing a black woman fit larger
cultural narratives, making it almost inevitable that third parties would insert
themselves into the struggle over media framing.
117
The use of this high-visibility
case as a platform for advancing more-general political agendas is also nothing
unusual and, leaving aside the particulars of the case, is not even undesirable.
After all, we are constantly being told that the United States enjoys a vigorous
marketplace of ideas,
118
so there is nothing wrong with spokespersons for
various ideological points of view seizing on big news stories to promote their
political interpretations and preferences. What was obviously problematic in
this case was that these spokespersons tended to treat the defendants as guilty
and used that conclusion as a basis for their larger political arguments about
white racism and privilege. The weak factual basis of this particular case does
not in itself logically preclude the possibility that their arguments to the more
114. See BENNETT ET AL., supra note 3, at 31–33 (discussing the failure of the Democratic party to
successfully promote a competing frame in the aftermath of 9/11); Entman et al., supra note 3, at 13–15.
115. On the stickiness of initial event schemas, see Robert Entman, Framing U.S. Coverage of
International News: Contrasts in Narratives of the KAL and Iran Air Incidents, 41 J.
COMM., Autumn
1991, at 6.
116. Furthermore, several stories reinforcing Nifong’s credibility would seem to undermine attempts
to promote this frame. Early stories describe Nifong as an able prosecutor, and describe his local
reputation as good. Ironically, an April 12, 2006, New York Times story dealing with the case and the
election features defense lawyers essentially agreeing with that characterization, perhaps bolstering
Nifong’s credibility. See Juliet Macur & Duff Wilson, Duke Inquiry to Continue, and So Will a
Campaign, N.Y.
TIMES, Apr. 12, 2006, at A14.
117. In this case, those third parties included black citizens in Durham and other groups that put
political pressure on the authorities, as well as the Duke faculty, who were concerned about the
university’s heavy-drinking culture and widespread flouting of alcohol laws, its emphasis on big-time
athletics, and the possible dilution of academic standards.
118. For research contesting this claim, see C. E
DWIN BAKER, MEDIA CONCENTRATION AND
D
EMOCRACY: WHY OWNERSHIP MATTERS (2007); BENNETT ET AL., supra note 3; Gadi Wolfsfeld &
Tamir Sheafer, Competing Actors and the Construction of Political News: The Contest over Waves in
Israel, 23 P
OL. COMM. 333 (2006).
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
Autumn 2008] RACE TO JUDGMENT 129
general points were valid. A large body of evidence supports the view that
white privilege and white racism do continue to oppress African Americans,
particularly poorer African Americans.
119
And there is a sound basis in this
particular instance for suggesting that Duke University is a relative bastion of
white privilege in a city where many working-class whites, blacks, and Latinos
struggle to make ends meet and to receive adequate educations, job
opportunities, health care, and other privileges that most Duke students (and
professors) take for granted. In the context of increasing income- and wealth
inequality nationally, and of continued black–white disparities across an array
of indicators,
120
the Duke lacrosse story, as it first seemed to be, was a
reasonably compelling symbol.
The very ability of a conference held at a prestigious law school to command
attention—and attendance—from important elites testifies to the accuracy, at
least in part, of the stereotype of Duke University and its students as unusually
privileged.
121
When Richard Jewell was hounded for his alleged, and ultimately
disproved, responsibility for the Atlanta Olympic bombing, no serious books
were written, the media paid minimal attention to the injustices, and no major
law school devoted a conference and law-review issue to the lessons from that
case. Richard Jewell was a working-class white man, a point that suggests part
of the Duke lacrosse case narrative does indeed properly concern class
privilege.
122
In many ways the Duke lacrosse story—with its simplifications, mistakes,
distortions, and initial pro-prosecution slant—was typical. Yet the case was also
quite unusual for crime coverage. Most crime news depicts poor, young whites
and nonwhites unaffiliated with prestigious institutions, and much of it
reinforces antiblack stereotypes.
123
This unusual case of white students at an
elite university subjected to the same kind of treatment customarily bestowed
on poor and working-class whites and minorities could yield social benefits. It
could raise awareness among journalists and law-enforcement officials of the
119. See, e.g., MICHAEL K. BROWN, ET AL., WHITEWASHING RACE: THE MYTH OF A COLOR
BLIND SOCIETY (2003).
120. See generally N
ATIONAL URBAN LEAGUE, ANNUAL REPORT ON SOCIO-ECONOMIC
CONDITIONS IN BLACK AMERICA, http://www.nul.org/thestateofblackamerica.html (discussing
multiple markers of black socioeconomics); D
AVID CAY JOHNSTON, FREE LUNCH: HOW THE
W
EALTHIEST AMERICANS ENRICH THEMSELVES AT GOVERNMENT EXPENSE (AND STICK YOU
WITH THE BILL) (2007) (discussing growing income inequality).
121. Disclosure: Entman and his wife are Duke alumni, and their son is a member of the class of
2008.
122. Jewell died just before this law-school conference. Not only was he innocent, he was actually
the man whose discovery of the bomb helped minimize casualties from it. See Kevin Sack, Richard
Jewell, 44, Hero of Atlanta Attack, Dies, N.Y. T
IMES, Aug. 30, 2007, at C14. See generally Kevin Sack,
He Felt Much ‘Like a Hunted Animal,’ N.Y. T
IMES, Oct. 29, 1996, at A12. Only two law-review articles
(written by the same authors) were located that centrally concern Jewell. See Clay Calvert & Robert D.
Richards, A Pyrrhic Press Victory: Why Holding Richard Jewell is a Public Figure is Wrong and Harms
Journalism, 22 L
OY. L.A. ENT. L. REV. 293 (2002); Clay Calvert & Robert D. Richards, Journalism,
Libel Law and a Reputation Tarnished: A Dialogue with Richard Jewell and His Attorney, L. Lin Wood,
35 M
CGEORGE L. REV. 1 (2004).
123. See discussion supra III.
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
130 LAW AND CONTEMPORARY PROBLEMS [Vol. 71:93
damage inflicted by the media’s standard operating procedures on the humans
subjected to a media circus and on the presumption of innocence that
undergirds the criminal justice system. Comparing this atypical instance to
normal crime news should also illuminate the ways racial minorities are
subjected to a kind of double jeopardy: they face the institutionalized, pro-
prosecution news bias likely applied in their specific case, and the antiminority
(especially antiblack) schemas perpetuated by so many other media
productions.
124
To whatever extent the Duke lacrosse case interfered with that presumption
and imposed psychological, financial, social, and other costs on the team
members and others, it is a travesty of justice. Yet it does not embody a general
tendency of media to slant news of white defendants more unfavorably than
that of black (or Latino) defendants. Nor does research support the
generalization that media treat black victims of alleged crimes more
sympathetically than white alleged victims; on the contrary.
125
Yet the Duke
lacrosse case presented a moment when attention was focused on the potential
problems that can arise in the administration of justice. One of the media’s
failings in this case was in not using this opportunity to further a more thorough
and ongoing discussion of prosecutorial misconduct or of normal journalistic
practice in covering crime.
Based on these findings and the prior research, we tentatively suggest some
remedies for suboptimal crime-news coverage, in full recognition that their
practical implementation and acceptance are problematic. First, with respect to
juries, judges could routinely include explicit warnings based on social-science
research that jurors, particularly white jurors considering African Americans
and other nonwhites, bring to their roles unconscious mental associations that
may prejudice deliberations. A judge’s directives could also make clear that this
is not a matter of accusing whites of being racist, and similar instructions should
apply to African American, Latino, and Asian jurors so they can be more-or-
less race neutral. The point is that everyone thinks through cultural schemas
that encourage unconscious mental associations. Most whites, for example, are
not outright racists; few think of themselves as racist and most seek rather to
prove to themselves that they are not. Research has shown the best way to
combat this is by explicitly telling people what might be happening in their
minds subconsciously.
126
Research suggests that awareness of this context can
moderate the effects of the unconscious, negative associations in
decisionmaking.
127
124. See ENTMAN & ROJECKI, supra note 14.
125. See, e.g., E
NTMAN & ROJECKI, supra note 18, at 81.
126. M
ENDELBERG, supra note 65.
127. See M
ENDELBERG, supra note 65, at 209–36 (discussing “psychological mechanisms that [can]
hinder or facilitate the impact of implicit and explicit racial appeals”); Devine, supra note 16, at 15
(discussing results showing that automatic prejudice activation occurs for both high- and low-prejudiced
subjects, but that controlled processing inhibits this for individuals who seek to maintain their
unprejudiced self identity); Kang, supra note 46, at 1528–35 (discussing arguments for and against the
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
Autumn 2008] RACE TO JUDGMENT 131
Another remedy would be training district attorneys, their staffs, police
departments, and others involved in the criminal-justice process to understand
the social science on unconscious racial stereotyping and decisionmaking.
Furthermore, training these personnel to recognize that their interactions with
the press may undermine the presumption of innocence could help mitigate
deleterious impacts on defendants’ rights. However, police officials, individual
officers, and certainly district attorneys have many political incentives to
promote one-sided publicity with the precise intention of prejudicing the jury
pool and improving the chances of conviction. The publicity also promotes an
image of competence and achievement that is simply good public relations for
police and prosecutors.
As illustrated in the Gentile case,
128
which dealt with the ability of the
defense to speak to the media, one of the problems for defense attorneys
appears to arise from American Bar Association standards discouraging both
sides from speaking to the press. Prosecutors’ and police officials’ incentives
and inclinations to violate the ABA’s rules appear strong. Defense attorneys on
the other hand seem more obedient, we speculate, perhaps because judges can
sanction them more readily than they can prosecutors. In addition, awareness of
a structural pro-prosecution bias rooted in the superior media resources of the
prosecution (which makes the playing field of frame contests uneven) could
discourage defense lawyers from even trying to contest the media frame. In a
media-saturated culture that includes 24/7 access to breaking news and to
unrestrained speculation on the Internet, it might make sense to revisit the
ABA guidelines.
These forces put more of an onus on journalists to alter their behavior. That
is, they should recognize the self-interested nature of public officials’
pronouncements and claims in criminal cases just as they appreciate—and often
report—the self-serving motives of public officials in other policy realms.
129
Journalists should also take into account the structural absence of a legitimate,
competing opposition of the sort that helps to promote (though certainly not
guarantee) more-balanced presentations of controversies in other policy areas.
130
As suggested by Smolkin, however, the forces that promoted the journalistic
shortcomings on exhibit in the Duke lacrosse case are not easily countered.
131
This leads to a reasonable prediction that prejudicial pretrial publicity and
troubling coverage associating minorities with crime and violence will continue
ability of context to overcome implicit racial attitudes); Lane et al., supra note 55, at 438–39 (discussing
research finding that exposure to counterstereotypical cues can reduce implicit racial bias).
128. Gentile v. State Bar of Nevada, 501 U.S. 1030 (1991). For a review and discussion of the ABA
guidelines with respect to trial publicity, see, for example, Ryan Brett Bell & Paula Odysseos, Sex,
Drugs and Court TV? How America’s Increasing Interest in Trial Publicity Impacts Our Lawyers and
the Legal System, 15 G
EO. J. LEGAL ETHICS 653 (2001).
129. See T
HOMAS E. PATTERSON, OUT OF ORDER (1994).
130. See B
ENNETT, supra note 113.
131. See Smolkin, supra note 94, at 20 (discussing the media’s inability to resist the “sensational,
simplistic storyline” in the Duke lacrosse case).
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
132 LAW AND CONTEMPORARY PROBLEMS [Vol. 71:93
to be the norm rather than the exception, ultimately causing more harm for
nonwhite than white defendants.
A
PPENDIX: CODING INFORMATION
For our analysis of the terms used in coverage, a research assistant searched
for the specific terms noted in the text. A research assistant would then
determine if the term was used in reference to what we were interested in, as
described in the text. Except where indicated, we accounted for the presence or
absence of terms at the level of the article as a whole and did not account for
the frequency with which terms were used within articles.
For the analysis of information that might promote in a neutral audience
member either a belief in the defendants’ guilt or in their innocence, a research
assistant coded each paragraph for the following:
1. The presence or absence of information that would tend to move a
neutral reader to be more likely to think that the events at the house
happened and that the players were guilty. Any information, assertions,
or statements that would tend to contribute to the belief that the event
happened and that the players were guilty was coded here as “guilty”
(including assertions or statements that the alleged acts happened or
may have happened, evidence or characterizations of evidence
consistent with the accuser’s claim, and claims that boosted the
credibility of the charges).
2. The presence or absence of information that would tend to move a
neutral reader to be more likely to think that events did not occur as the
accuser described or to be more likely to think that the players were
innocent. Any information, assertions, or statements that would tend to
contribute to the belief that this did not happen and that the players
were innocent was coded as “not guilty.” This included assertions that
the players were innocent and that the alleged acts did not happen;
evidence or characterizations of evidence consistent with claims that
they were not guilty, including inconsistency in the evidence; challenges
to the credibility of the accuser and challenges to the credibility of
Nifong; and mentions of Nifong’s errors, including critiques of his
behavior by the defense.
Every assertion about what happened, did not happen, or might have
happened was coded even when the assertion itself did not provide supporting
evidence for the claim. Thus, the statements of the woman who accused the
players of sexual assault were coded as supporting the claim of guilt, whereas
statements by the players or their lawyers that the defendants were innocent
were coded as “not guilty.” Assertions that contained conditional statements,
however, were not coded. In coding early stories (before specific individuals
were charged), any references to any lacrosse player that might move a neutral
reader toward a belief in guilt or innocence were coded. Character references
were coded only when they were used in the same paragraph as a reason for
05__ENTMAN & GROSS__CONTRACT PROOF_UPDATE.DOC 12/1/2008 3:14:01 PM
Autumn 2008] RACE TO JUDGMENT 133
believing in guilt or innocence, or when the character reference was
immediately preceded or followed by a paragraph making a link between the
character trait and likely guilt or innocence. Statements denying the accusations
that restate the accusations as a part of the sentence were not coded as
assertions of the truth of the accusations.
To ensure intercoder reliability, we adopted a strict coding standard in
which coders used only the information within the paragraph and the
immediately preceding paragraph in making their coding determination. This
can result in a conservative coding of the coverage. Information in the Duke
lacrosse case is certainly read in the context of an entire article or in the context
of what is known about the case at a given point; however, a strict coding
scheme facilitates reliability in systematic content analysis.
A second coder coded twenty percent of the total articles coded for the
analysis by paragraph in order to assess intercoder reliability. Intercoder
reliability was .87 using Scott’s Pi, a chance-corrected measure of reliability.
This provides a conservative measurement of intercoder reliability because
every instance in which one coder gauged the paragraph as both guilty and not
guilty and a second coder gauged it as only guilty was treated as an error (even
though in such a case the coders did agree on the notion that the paragraph
reinforced guilt, but disagreed on the notion that the paragraph also reinforced
innocence).
