Recent Police Killings in the United States: A Three-City Comparison
Angela S. Lee1, Ronald Weitzer1, and
Daniel E. Martı´nez2
Ronald Weitzer, Department of Sociology, George Washington University, Washington, DC 20052,
2018, Vol. 21(2) 196–222
! The Author(s) 2017
Reprints and permissions:
Task Force,’’ 2015; Wilson, 2015).
Recent police killings of citizens in the United States have attracted massive coverage
in the media, large-scale public protests, and demands for reform of police departments
throughout the country. This study is based on a content analysis of newspaper
coverage of recent high-profile incidents that resulted in a citizen’s death in
Ferguson, North Charleston, and Baltimore. We identify both incident-specific content
as well as more general patterns that transcend the three cases. News media
coverage of similar incidents in past decades tended to be episodic and favored the
police perspective. Our findings point to some important departures from this paradigm.
Reporting in our three cases was more likely to draw connections between
discrete incidents, to attach blame to the police, and to raise questions about the
systemic causes of police misconduct. These findings may be corroborated in future
studies of news media representations of high-profile policing incidents elsewhere.
Recent highly publicized cases of police misconduct in the United States have
catalyzed street demonstrations throughout the country, the Black Lives Matter
movement, the Blue Lives Matter countermovement, a presidential commission
on policing, and reform initiatives in several cities (Condon, 2015; ‘‘President’s
1Department of Sociology, George Washington University, Washington, DC, USA
2School of Sociology, University of Arizona, Tucson, AZ, USA
News coverage of police killings has increased
as well, partly because of the advent of video recordings of such events, public
protests after such incidents, and heated discussions on social media. The result
is that police misconduct has experienced an unprecedented ‘‘new visibility’’
(Goldsmith, 2010; Victor & McPhate, 2016).
Researchers have analyzed news media representations of policing incidents
that occurred decades ago, but little is known about the nature of such representations
today. News consumption has evolved over the years, with cable
television and social media helping to disseminate information much more
widely and immediately than in the past. Video coverage of police actions has
increased as well.1 Because of these developments, news media coverage of incidents
involving the police may be having a larger impact on public perceptions
and official responses than reporting of similar events in the past. Moreover,
publicized incidents of police misconduct can damage the reputation of police
not only in the city where an incident occurs but also nationwide, and this is
especially true when multiple events cluster in a compressed timespan—‘‘when
one dramatic incident occurs shortly after another [and] bears strong resemblance
to another case’’ (Lawrence, 2000, p. 103). It has been argued that
‘‘this contamination-by-association is occurring today in a cumulative manner—
with each incident pollinating subsequent ones—in part because activists
and the media are drawing connections between them’’ (Weitzer, 2015, p. 475). A
2014 poll reported that a sizeable minority of Americans (43%) believed that the
police killings of Michael Brown (in Ferguson) and Eric Garner (in New York)
were not ‘‘isolated incidents’’ but instead ‘‘a sign of broader problems in the
treatment of African Americans by police’’ (Washington Post/ABC News, 2014).
Two years later, the proportion taking the ‘‘broader problems’’ view had grown
to 60% (54% of Whites, 79% of Blacks), arguably because of an accumulation
of publicized incidents since the Brown and Garner killings in 2014 (Pew
Research Center, 2017).
The impact of news media representations on public perceptions is important.
Since most people have limited direct contact with police officers, information
about the police comes largely from the media, including traditional news
sources whose reporting is now often redistributed through social media (Pew
Research Center, 2016). The public does not necessarily adopt the news media’s
version of reality, but by setting the agenda for what is defined as news and
selectively presenting content, the news media strongly influences public perceptions
of events and issues (Iyengar & Kinder, 1987; McCombs & Shaw, 1972).
The pivotal role of the media has been demonstrated in studies documenting
enhanced support for the police among people who watch reality-TV programs,
such as Cops, that present officers in a sympathetic light (Eschholz, Blackwell,
Gertz, & Chiricos, 2002) as well as erosion of public confidence in the police after
well-publicized incidents of police misconduct (Kaminski & Jefferis, 1998;
Sigelman, Welch, Bledsoe, & Combs, 1997). And these outcomes are especially
Lee et al. 197
likely for people who are frequently exposed to media representations of the
police. One survey found that the more a citizen read newspaper accounts of an
incident involving drunk Indianapolis police officers who beat two citizens, the
greater the likelihood that these readers would deem the officers guilty
(Chermak, McGarrell, & Gruenewald, 2006). Similarly, a national study
found an association between an individual’s perceptions of the police and his
or her exposure to news of police misconduct: Individuals heavily exposed to
such reporting were more likely to view several types of police misconduct as
widespread and to endorse a host of reforms in policing (Weitzer & Tuch, 2006).
News coverage, especially when it includes video recordings or public protests
after an incident, can also affect police officers themselves. Officers typically
rationalize and justify their use of force, whether excessive or not (Waegel,
1984; Weisburd, Greenspan, Hamilton, Bryant, & Williams, 2001), but they
may also alter their behavior in response to public criticism of the police or if
they feel they are being monitored. Officers may avoid certain types of encounters
altogether or may consciously temper their treatment of the citizens they
interact with (Simonson, 2016). A few studies document the latter. Half of the
officers interviewed in a recent Canadian study said that, because of the potential
for video recording by citizens, they now use force less often and use a lesser
amount of force in specific encounters; and three quarters reported other behavioral
changes for fear of being caught on camera (Brown, 2016). In Britain, two
thirds of the officers interviewed in six towns said that the presence of CCTV
cameras in public places made them ‘‘more careful’’ in conforming to procedural
requirements while on patrol and anxious about their conduct being scrutinized
after the fact; some officers stated that the cameras made them more reluctant to
use force against citizens (Goold, 2003). Regarding the effect of controversial
incidents, a survey of 7,917 police officers across the United States reported that
the vast majority of officers believe that the recent fatal encounters with citizens
and the public outcry generated by them has made their job harder (86%), that
officers are now more concerned about their safety (93%), and that officers are
less willing to stop and question suspicious people or to use force when it is
called for (72%; Pew Research Center, 2017). At the same time, half of these
officers say that wearing body cameras will make officers act more appropriately
when dealing with the public. This view received empirical support in randomized
field experiments in Rialto, CA, and Orlando, FL, which found that
equipping officers with body cameras correlated with reductions in the use of
force as well as decreased complaints from the public (Ariel, Farrar, &
Sutherland, 2014; Jennings, Lynch, & Fridell, 2015).
News media representations of the police are therefore important in multiple
ways: They can influence public perceptions, catalyze popular demands for
reform, affect the conduct of at least some officers, and help generate initiatives
to curb police misconduct. But these outcomes depend in part on the nature of
news media coverage. This article examines reporting on high-profile events in
198 Police Quarterly 21(2)
three cities, identifying (a) the main thematic similarities across the cities and (b)
the issues on which reporting varies by city, suggesting local contextual explanations
for the incident.
News Media Constructions of Policing Issues
Decades of research shows that political elites are the primary definers of events
and issues covered in the mass media. Journalists typically perceive government
officials as the most credible authority in their domain and rely heavily on them
as sources (Bennett, 1996; Cook, 1998). This is especially the case in crime and
justice reporting, where police officials typically have a monopoly on key information.
Certain reporters are assigned to cover local crime stories—the ‘‘crime
beat’’—and develop symbiotic relationships with police officials. Some reporters
come to identify with police values and may thus distort news stories in favor of
the police (Chermak & Weiss 2005; Ericson, 1995; Lawrence, 2000). At the same
time, journalists tend to marginalize the voices of nonofficials due to the latter’s
lack of demonstrable ‘‘credibility’’ or lack of access to reporters (Lawrence,
2000, p. 55; Paletz & Entman, 1981).
News media coverage rarely focuses on larger patterns or causes of police
misconduct (Lawrence, 2000; Pollack & Allern, 2014). Instead, reporting tends
to be event oriented and fragmentary:
The news doesn’t often address questions such as the prevalence of police violence,
patterns in how it occurs, or the acceptability of police tactics in fighting crime and
maintaining order . . . Not only is most news about police use of force highly episodic
in its focus. Most use-of-force incidents that are reported in the news disappear
from the news pages quickly. (Lawrence, 2000, p. 45)
Likewise, the causes of police brutality are typically individualized—either blaming
a ‘‘rogue cop’’ or the citizen (for provoking the officer)—rather than being
defined as outcomes of systemic problems in a police department or larger societal
conditions (Lawrence, 2000; Pollack & Allern, 2014). Finally, media reporting
on the police is influenced by the core imperative of crime control: Officers
are involved in dangerous work and sometimes have to make split-second decisions,
which may give them the benefit of the doubt in ambiguous situations.
Therefore, it is not just the fact that journalists are occupationally dependent on
official sources, but also the widespread fear of crime and diffuse public support
for punitive responses that helps to privilege the police version of events when
covered by the news media.
Some scholars argue that skewed media coverage of police killings promotes
the normalization of police violence. In her content analysis of New York Times
and Los Angeles Times articles published between 1985 and 1994, Regina
Lawrence (2000) found that police use of force was ‘‘typically normalized in
Lee et al. 199
news that presents it as a necessary and appropriate response to a violent world’’
(p. 60). Similarly, fully 70% of a sample of 105 newspaper articles on police use
of deadly force, published between 1997 and 2000, provided justifications for the
killings and presented officers as acting reasonably under the circumstances
(Hirschfield & Simon, 2010). Of course, news stories in the past occasionally
departed from this narrative, particularly if a video recording or other compelling
evidence existed. In these instances, ‘‘a sensational police killing can shift
patterns of symbolic construction in police violence news, even if only temporarily’’
(Hirschfield & Simon 2010, p. 156). In these cases, police conduct is
reframed as problematic and the media may give voice to nonofficial sources.
Examples of this include the Rodney King beating in Los Angeles in 1991 and
the killing of Amadou Diallo in New York in 1999 (Hirschfield & Simon, 2010;
Lawrence, 2000). But this kind of critical coverage is rare, according to the
We know of no research on how the news media have portrayed recent
instances of police violence. The current study explores this question by examining
local newspaper coverage of three recent highly publicized killings: Michael
Brown in Ferguson, Missouri; Walter Scott in North Charleston, South
Carolina; and Freddie Gray in Baltimore, Maryland. Each incident received
massive coverage by both local and national news organizations. Two research
questions drive the study: First, what are the main types of content appearing in
newspaper coverage of these incidents across the three cities? Do the patterns of
episodic coverage, individualization, and normalization—documented in earlier
studies—still predominate today? Second, on which content issues does coverage
vary by city? The former question allows us to identify policing issues that
transcend city context and may reflect broader constructions of policing problems
throughout the nation, whereas the latter question allows us to contextualize
what appear to be city-specific issues that shape each event.
The Three Incidents and Cities
Michael Brown, an 18-year-old unarmed Black man, was shot and killed by a
White police officer, Darren Wilson, during a confrontation in Ferguson, MO,
on August 9, 2014. The incident was not video recorded, but the authorities
subsequently released video footage suggesting that Brown had engaged in a
robbery at a convenience store shortly before being stopped by Officer Wilson
for walking in the center of a street. Days of street protests and rioting followed
the initial media reporting on the incident, events that showed a heavily militarized
police force that was later augmented by the National Guard. A grand jury
decided not to indict Wilson, and the Justice Department conducted its own
investigation in March 2015, clearing Wilson of civil-rights violations.
The second incident occurred in North Charleston, SC. On April 4, 2015, a
White officer, Michael Slager, stopped a 50-year-old Black man, Walter Scott,
200 Police Quarterly 21(2)
because of his car’s faulty brake light. During the encounter, Scott ran away and
shortly thereafter a physical altercation ensued between the two men,
which included struggling over the officer’s Taser. As Scott fled, Slager
fired his handgun eight times, hitting Scott in the back five times. A bystander’s
recorded the shooting on his cell phone—video that contradicted Slager’s
account of the shooting and resulted in him being fired and tried for murder.
A mistrial was declared after the jury deadlocked in the state trial in late 2016. In
the subsequent federal trial in May 2017, Slager pleaded guilty to a civil rights
charge and currently awaits sentencing.
On April 12, 2015, just 8 days after Scott’s shooting, Freddie Gray was
arrested in Baltimore, MD, for allegedly possessing an illegal knife. A bystander
video recorded part of Gray’s encounter with six officers, showing him being
dragged to a police van and having difficulty standing on his own. The officers
failed to provide medical attention to Gray, who had asthma, after he requested
an inhaler and appeared to have trouble breathing (Cohn, 2015). The officers
failed to secure Gray inside the van, and reports in the media speculated that the
driver may have deliberately given Gray an injurious ‘‘rough ride,’’ something
Baltimore police had been accused of doing in the past. During the ride, Gray
fell into a coma and died a week later. After the medical examiner ruled Gray’s
death a homicide, the State’s Attorney for Baltimore filed charges against the six
officers—charges that include illegal arrest, assault, and second-degree murder.
Four of the officers were prosecuted but not convicted, and the two remaining
cases were dropped by the prosecution in July 2016. One month later, the
Justice Department issued a scathing report on Baltimore’s police department,
which detailed patterns of excessive force and racially biased policing
(Department of Justice [DOJ], 2016). Video cameras have now been installed
in the city’s police vans.
To contextualize our findings, Table 1 provides selected demographics on the
three cities and their police departments. African Americans have substantially
higher poverty and unemployment rates than Whites, roughly double or triple in
each city. The cities also have plurality- or majority-Black populations; yet, the
police departments in Ferguson and North Charleston were overwhelmingly
composed of White officers (83% and 81%, respectively), while Baltimore’s
was more mixed, with a slight majority of White officers. At the time of the
incident, the police chiefs in Ferguson and North Charleston were White
and Baltimore’s was African American. All three police chiefs have now
Data and Methods
This study is based on a content analysis of newspaper articles from the cities
where the events occurred. Therefore, the units of analysis are newspaper articles.
The articles consist of straight news reporting, editorials, op-eds, and
Lee et al. 201
Associated Press articles published in these newspapers. The three incidents were
selected because they involved a highly publicized police killing of a citizen,
occurred within a fairly narrow time span (8 months), and garnered substantial
local and national news coverage.2 Our sources were the St. Louis Post-Dispatch
for Michael Brown in Ferguson, the Post and Courier for Walter Scott in North
Charleston, and the Baltimore Sun for Freddie Gray in Baltimore. The timespan
of the articles used for this research begins on August 9, 2014 (the day after
Brown’s death), and concludes on September 30, 2015. A total of 578 articles
were collected: 267 from the St. Louis Post-Dispatch, 119 from the Post and
Courier, and 192 from the Baltimore Sun. The fact that the three cases generated
this much coverage at the local level is noteworthy in itself.
Lexis-Nexis was used to collect articles from the Baltimore Sun and St. Louis
Post-Dispatch and, because Lexis-Nexis does not contain Post and Courier
articles, they were accessed through an online subscription. Key search terms
consisted of ‘‘police’’ and the full name of the deceased person (hereafter,
‘‘victim’’). To capture all articles pertaining to the incidents, a search limited
Table 1. Selected City Characteristics.
Population 21,151 102,143 622,271
Whites 31% 38% 30%
Blacks 66% 47% 63%
Hispanics 1% 11% 5%
Police department composition
Whites 83% 81% 51%
Blacks 11% 16% 40%
Hispanics 4% 3% 7%
Whites 9.7% 15.8% 14.8%
Blacks 27.4% 32.4% 28.3%
Hispanics 13.8% 30.3% 25.5%
Whites 6.5% 7.8% 7.1%
Blacks 15.7% 16.4% 18.5%
Hispanics 2.9% 7.7% 9.7%
Sources: U.S. Census, 2014 American Community Survey. Figures on police department
composition are from Bureau of Justice Statistics, Law Enforcement
Management and Administrative Statistics, 2013 survey, reported in Governing,
202 Police Quarterly 21(2)
to the full name of the victim (without ‘‘police’’) was also conducted in order to
include other relevant articles. Finally, because the names of the victims may not
have been available for the initial articles, a search using only ‘‘police’’ was
conducted for the first 2 weeks after each incident was first reported. All articles
were sorted by date to eliminate duplicates.
Once the articles were collected, we constructed a list of major codes and
subcodes, facilitated with the qualitative software program Atlas/ti.3 These are
grounded, inductive codes that emerged from the narratives, not superimposed a
priori. Major codes are defined as types or categories of content (e.g., ‘‘Causes of
Incident’’) and subcodes as subsidiary items under the rubric of a major code
(e.g., ‘‘Victim Precipitation’’; see Strauss & Corbin, 1990). We recorded the
presence or absence of codable content in each article and allowed for multiple
mentions of subcodes within each major code. Codes were tagged to each article
if they were mentioned (yes¼1) at any time within the article; all other codes
were marked as ‘‘not mentioned’’ (no¼0). Additionally, close attention was
given to specific aspects of each newspaper’s coverage. Coding yielded 7 major
codes and 29 subcodes.
To assess the reliability of the coding process, we drew a 10% random sample
of St. Louis Post-Dispatch articles (N¼27). We then enlisted a criminology
graduate student, trained in the initial coding scheme, to code the randomly
selected articles, using the major codes and subcodes that emerged from the
authors’ coding procedure. Intercoder agreement occurred in 99% of the
coded observations (N¼972) in the subset of articles. The amount of agreement
is likely due to the rigorous training the second coder underwent.
The analytic plan was twofold. We used a univariate (i.e., descriptive) approach
to identify the prevalence of coded content both within and across all three incidents.
We then conducted Pearson’s chi-square tests to examine the bivariate
relationships between the codes and the specific incidents. The Pearson’s chisquare
tests allowed us to identify statistically significant differences between the
prevalence of major codes and subcodes for the three incidents.
The newspapers routinely identified the victims in Ferguson and North Charleston
as unarmed Black males and the officers involved as White males. For Baltimore,
media reporting varied in whether it characterized Freddie Gray as being ‘‘armed’’
(a knife in his pocket) and whether it mentioned the mixed racial makeup of the
six officers involved (three White and three Black). In addition to these baseline
patterns, our major codes helped us identify the types of content or topics mentioned
frequently in newspaper coverage across the three incidents. Given the
different circumstances of the three cases, these overarching themes point to
issues that transcend city context and may reflect broader constructions of the
problem of police misconduct throughout the nation.
Lee et al. 203
Table 2. Major Codes and Subcodes by Total Sample and City.
(N¼578) (N¼267) (N¼119) (N¼192) level
Causes of incident 27.7 25.5 44.5 20.3 ***
15.4 9.4 31.1 14.1 ***
Racism 5.7 7.5 7.6 2.1 *
Rotten apple 5.7 6.4 10.0 2.1 **
2.8 1.9 1.7 4.7 NS
Police policies 2.0 1.1 5.0 1.6 *
1.7 1.9 2.5 1.0 NS
Lack of diversity 0.9 1.5 0.8 0.0 NS
Lack of officer
0.9 0.8 0.8 1.0 NS
36.5 31.1 52.1 34.4 ***
Blaming police 32.5 25.1 50.4 31.8 ***
Blaming victim 14.5 20.2 10.1 9.4 **
26.0 25.8 9.2 36.5 ***
Police issues 54.2 50.6 53.8 59.4 NS
Police violence 40.0 33.3 41.2 48.4 **
16.6 18.4 13.5 16.2 NS
15.0 12.4 23.5 13.5 *
10.0 9.4 8.4 12.0 NS
Policing is hard job 3.6 3.8 3.4 3.7 NS
Racial issues 42.6 51.3 44.5 29.2 ***
Institutional racism 17.8 19.9 19.3 14.1 NS
Racial profiling 17.5 20.6 22.7 9.9 **
Racism (other) 16.8 29.2 2.5 8.3 ***
Racial disparity 13.7 16.1 19.3 6.8 **
204 Police Quarterly 21(2)
Table 2 displays proportions for coded content for the total sample as well as
for each incident covered by the three newspapers.4 Univariate and bivariate
results are presented later, along with illustrative quotations from the articles.
We organized our seven major codes under two rubrics: Incident-Specific Issues
and Structural Factors.
Studies have shown that newspaper coverage of police violence in the past has
overwhelmingly centered on the specific incident per se. The current study found
that much of the recent coverage fits this pattern. The rubric Incident-Specific
Table 2. (continued)
(N¼578) (N¼267) (N¼119) (N¼192) level
Denial that racism
1.6 2.3 1.7 0.5 NS
15.4 21.4 8.4 11.5 ***
Reforms 46.2 52.4 61.3 28.1 ***
Body cameras 21.1 15.4 51.3 10.4 ***
11.8 22.5 4.2 1.6 ***
10.0 12.0 5.0 10.4 NS
9.7 13.5 4.2 7.8 **
9.2 13.9 8.4 3.1 ***
5.5 8.6 2.5 3.1 *
4.0 6.0 0.8 3.1 *
Decarceration 2.2 1.9 0.8 3.7 NS
1.4 0.8 0.0 3.1 *
Note. NS¼Not statistically significant.
*p<.05. **p<.01. ***p<.001.
Lee et al. 205
Issues consists of two major types of content: Causes of Incident and
Responsibility for Incident.
Causes of incident. As briefly mentioned earlier, more than one quarter (27.7%) of
articles mentioned Causes of the Incident, with notable variation in the appearance
of this major code across the three incidents: 25.5% St. Louis Post-
Dispatch, 44.5% Post and Courier, and 20.3% Baltimore Sun (p<.001). Eight
subcodes were identified as reasons for the killings, and three appeared with
notable frequency in the total sample: Victim Precipitation (15.4%), Racism
(5.7%), and Rotten Apple (5.7%).
Victim Precipitation is when a citizen allegedly behaved in a way that precipitated
or escalated the incident and resulted in the police officers using force
against them—thus taking the onus off the officer. Mentions of the victim resisting
the officer, attacking the officer, fleeing the scene, or brandishing a weapon
were coded as Victim Precipitation. Victim precipitation accounted for 15.4% of
the 578 articles. The St. Louis Post-Dispatch published articles that explained the
killing resulting from a physical altercation between Officer Wilson and Brown.
Articles also stated that after being shot, Brown’s body lurched forward as if he
were going to charge at Wilson. As for Post and Courier articles, nearly all
mentions of victim precipitation involved resistance by Scott: He initially fled
the scene, was caught and struggled with the officer, and fled a second time.
The Baltimore Sun reported that Gray ran when he saw one of the officers.
Victim precipitation as a contributor to the killing was mentioned most
frequently in Post and Courier articles on Scott (31.1%).
Racism was depicted as a cause of the incident in roughly 6% of all articles.
This theme emerged when the newspapers cited racism, either on the part of the
individual officer or entire police department, as an explanation for the incident.
Some racism content appeared bluntly, such as a woman who declared, ‘‘Racism
killed Brown’’ (quoted in Deere, 2014, p. A1). Another article quoted the president
of North Carolina’s National Association for the Advancement of Colored
People chapter, William Barber: ‘‘All of these things should force us to recognize
that we still have systemic racism’’; the op-ed author agreed and added that the
Scott incident ‘‘proved indisputably that racial profiling is very real, that in
America a Black man can be killed for having a faulty brake light on his car’’
The Rotten Apple subcode refers to the police officers involved in an incident
as aberrant within the police department. The cause of the incident was due to
the officer’s alleged character or bias. This subcode was present in nearly 6%
of all articles. In the St. Louis Post-Dispatch, Officer Wilson was portrayed as
callous and unprofessional. One article reported that Wilson ‘‘pulled up and
ordered [Brown and his friend] to ‘get the F on the sidewalk’ and grabbed
Brown, 18, in the throat’’ (Kohler, 2014, p. A9). Referring to Wilson, Police
Chief Frank McCall declared: ‘‘You’ve come across a bad apple . . . There are
206 Police Quarterly 21(2)
methods and ways of going about getting that bad apple plucked out’’ (Bock,
2014, p. A11). The Rotten Apple theme appeared in a higher proportion of
Post and Courier articles (10%) than in the other two newspapers, probably
because of the bystander’s video of Scott’s shooting and the subsequent
murder charges leveled against Officer Slager, whose actions were labeled
those of a rogue cop.
Responsibility for incident. Approximately 37% of the total sample contained content
describing Responsibility for the Incident. This is an important theme, as it
reveals whose side, if any, the newspapers’ accounts supported—the police officer’s
or victim’s. Studies cited earlier have found that the news media relies
heavily on the authorities as sources and skews reporting in their favor. Does
coverage of recent police shootings, in our three cities, bear this out?
We found that the police were blamed more than twice as often as the citizen
(32.5% vs. 14.5%). North Charleston Mayor Keith Summey stated that Officer
Slager had made a ‘‘bad decision’’: ‘‘When you’re wrong, you’re wrong. If you
make a bad decision, don’t care if you’re behind the shield or just a citizen on the
street, you have to live by that decision’’ (quoted in Knapp, 2015c). This statement
was quoted in several articles in the Post and Courier. A Ferguson example
is an editorial critical of Officer Wilson: ‘‘We’ve got some John Wayne cops who
don’t know that ‘tactical withdrawal’ is an honored military strategy’’ (‘‘Partisan
melee over law,’’ p. A14). When the Baltimore Sun attached blame to the officers,
it typically described their behavior as abusive or insensitive. Consider this
The video shows . . . Gray screaming on the ground with police kneeling beside him
before he’s dragged to the police van, where he appears to stand briefly. Witnesses
have said Gray’s legs looked broken and suggested the injury may have occurred
during his arrest. Police acknowledged Gray was having trouble breathing and
asking for an inhaler for asthma. Police now say he should have received medical
treatment before being loaded into the van, where they also say they failed to
buckle him in. (Cohn, 2015, p. A14)
And an editorial unequivocally contested the rationale for arresting Gray:
‘‘Officers had no probable cause to chase Gray when he ran after making eye
contact with one of them, and no probable cause to restrain, search, and arrest
him’’ (‘‘What Took So Long?,’’ 2015, p. A20).
Although video footage captured part of the Gray incident—Gray being
dragged to a police van—it did not reveal what happened inside the van. That
a video documented Scott’s shooting may explain why the blaming-police theme
is reflected in a much higher proportion of Post and Courier articles (50.4%) than
in the St. Louis Post-Dispatch (25.1%) or Baltimore Sun (31.8%). That video
evidence can speak volumes was also clear in the October 2014 killing of Laquan
Lee et al. 207
McDonald in Chicago: The officer involved was charged with murder only after
a video of the shooting was released a year after the killing.
Fewer articles contained content coded Blaming Victim. The St. Louis Post-
Dispatch reported not only that Brown struggled with the officer prior to being
shot but also frequently mentioned security-camera footage of Brown stealing
cigars from a convenience store and shoving the store attendant prior to his
encounter with Officer Wilson. The implication was that Brown was dangerous
and that Wilson’s actions may have been justified. This security footage itself
became a contested issue in the St. Louis Post-Dispatch because it was released to
the public at the same time as Wilson’s name, with some analysts surmising that
the timing was intended to defuse public outrage over the killing. And this may
account for the Blaming Victim code materializing twice as often in the St. Louis
Post-Dispatch than the other two newspapers. Victim blaming was evident in one
tenth of Post and Courier articles, for example, by reporting that Scott posed a
threat to or endangered the life of Slager. This angle was more evident in early
Post and Courier coverage, before the video of the shooting was released: ‘‘Police
allege that during the struggle the man gained control of the Taser and
attempted to use it against the officer. The officer then resorted to his service
weapon and shot him’’ (Elmore & MacDougall, 2015). Articles in the Baltimore
Sun mentioned Gray’s possession of a knife, which police alleged to be illegal,
Gray’s testing positive for opiates and cannabis at the time of his arrest, and
Gray’s previous schemes to injure himself while in police custody in order to
collect settlement money.
The final theme is Questioning the Investigation. Unlike previous research
that found the media rarely interrogating the authorities’ investigation of a
police-involved incident, our three newspapers were not averse to doing so.
The mean (26%) is skewed by the low percentage of mentions of this theme in
North Charleston (9.2%), arguably because the officer involved was immediately
charged with murder once the videotape surfaced. Investigation of the incident
was questioned much more often in Ferguson (25.8%) and Baltimore (36.5%)
because of intense controversy in Ferguson over the much-delayed naming of the
officer and the decision not to indict the officer, and in Baltimore, the delay in the
decision to indict the six officers. Delays of any kind or failure to prosecute can
be interpreted as signs of flawed investigations.
Structural factors. It is not surprising that Incident-Specific Issues would figure
prominently in our data, especially in light of previous research findings that
media coverage of instances of police misconduct has been heavily skewed in
focusing on particular events rather than larger systemic factors. Our analysis,
however, reveals a clear departure from past media coverage in that several
important Structural Factors were highlighted in the three newspapers. Here,
structural refers to larger institutional or societal factors that were presented as
underlying or related to the incidents. And, considering the different
208 Police Quarterly 21(2)
circumstances of the three incidents, these overarching themes point to issues
that transcend city context and may reflect broader constructions of the problem
of police misconduct throughout the nation. We identified four major codes that
constitute Structural Factors: Police Issues, Racial Issues, Poverty/Inequality,
and Reforms. With the exception of Police Issues, there was notable variation in
mentions of these main themes across the three incidents, which we discuss in the
A large number of articles discussed general policing issues beyond the particular
incident. In fact, 54.2% of the articles in the total sample were coded Police
Issues. The most frequently mentioned subcodes were Police Violence, Police-
Community Relations Strained, and Accountability Deficiency.
Police Violence (40%) refers to any mention of police violence outside of our
three incidents. It appeared in articles mentioning specific, prior incidents of
police violence in any city as well as in general critiques regarding use-of-force
actions. Mentions of Police Violence usually portrayed the police in a negative
light. Some St. Louis Post-Dispatch articles drew a connection between race and
police violence or compared Brown’s death with police killings of other individuals.
Post and Courier articles ranged from mentioning police treatment of civil
rights protesters on ‘‘Bloody Sunday’’ in Selma, Alabama, in 1965 to Officer
Slager’s unlawful Tasing of a man in a prior encounter. Mentions of police
violence also included general statements, such as when the Post and Courier
reported that the Scott incident ‘‘renewed scrutiny of police officers’ use-of-force
against Black men and ignited challenges to patrolling methods that had long
been criticized in North Charleston’’ (Knapp, 2015a). In addition to citing other
use-of-force incidents and making generalizations regarding police violence
toward African Americans, Baltimore Sun articles revealed the Baltimore
police department’s history of giving ‘‘rough rides’’ when transporting a suspect.
Such individuals cannot protect themselves if they are handcuffed but unsecured
and thrown around inside a police vehicle that is driven aggressively. An article
in the Baltimore Sun defined a ‘‘rough ride’’ as when ‘‘officers intentionally drive
erratically, causing shackled passengers to bounce helplessly against the walls of
the van’’ (Marbella, 2015, p. A1). As the Baltimore Sun speculated about
how Gray sustained injuries in the van, the newspaper also drew connections
to a larger pattern involving rough rides that had come to light in Baltimore in
Police-Community Relations Strained appeared in references to citizens’ preexisting
distrust in the police or a worsening of police-community relations
because of the incident in question. This code accounted for 16.6% of articles
in the total sample (18% St. Louis Dispatch-Post; 13% Post and Courier; 16%
Baltimore Sun). Examples include a ‘‘Improved dialogue on police’’ (2015)
Lee et al. 209
editorial on Scott’s killing: ‘‘It has shaken the community’s confidence in law
enforcement,’’ and an article highlighting ‘‘strained relations between the local
police force and Black community members’’ stemming from this killing
The Accountability Deficiency code refers to content noting that the police
are rarely held accountable for their misconduct generally—apart from the three
incidents examined here. Accountability is mentioned in 15% of all articles, and
almost one quarter in North Charleston. Two examples illustrate this theme:
Too many times communities across the country have watched officers exonerated
in controversial shootings . . . Every time they wound communities, and those scars
never really heal. (Hicks, 2015b)
Baltimore has paid more than $6 million in judgments and settlements in civil suits
against police during the past five years . . . Police are rarely charged after encounters
that result in deaths and are even more rarely convicted as a result. (‘‘Mosby’s
‘Conflicts’,’’ 2015, p. A1)
Fully 43% of articles in the total sample mentioned broad or generic racial issues
in their coverage of the incidents: 51.3% St. Louis Dispatch-Post, 44.5% Post
and Courier, and 29.2% Baltimore Sun (p<.001). Notable subcodes in the Racial
Issues category include Institutional Racism (17.8%), Racial Profiling (17.5%),
and Racial Disparity (13.7%). The relatively high frequency of mentions suggests
that newspapers are beginning to draw connections to racial issues beyond
the specific incident—a departure from media coverage of police misconduct in
Institutional Racism implies that racism is structurally embedded in a police
department. In the Post and Courier, for example, this code appeared in phrases
such as ‘‘stop racist police terror’’ (Knapp & Smith, 2015) and calls ‘‘to change a
culture of fear-induced bigotry in and beyond law enforcement’’ (Darby, 2015).
Institutional Racism was mentioned in nearly 18% of articles in the total sample,
and there were no notable differences in the proportions of mentions between the
three incidents, suggesting that this issue transcends the three incidents.
Racial Profiling (17.5%) refers to the practice of police officers stopping
people based on their race. There were notable variations in the appearance of
this code across the three newspapers: 20.6% St. Louis Post-Dispatch, 22.7%
Post and Courier, and 9.9% Baltimore Sun (p<.01). The lower proportion of
articles in the Baltimore Sun may be due to the fact that three of the six officers
involved in the Gray incident were African American or perhaps to the higher
representation of Blacks on Baltimore’s police force (40%) compared with the
other two largely White police departments (see Table 1). In the other two cities,
210 Police Quarterly 21(2)
articles portrayed Brown and Scott as having been stopped solely for being
Black, which they connected to racial profiling more generally. An editorial
on Ferguson revealed that ‘‘Blacks were 37% more likely to be pulled over in
2013 than Whites, as a percentage of their respective populations’’ and that
Blacks were ‘‘twice as likely to be searched for contraband . . . even though
police found contraband, percentage-wise, more often in the cars of White drivers’’
(‘‘Justice Department investigation,’’ 2014b, p. A12). A Post and Courier
writer stated, ‘‘Some Black friends . . . have long and frequently told me about
being harassed by police officers without credible cause. Those pals have been
convinced of, and convincing in their assertions of, racial bias in those encounters’’
(Wooten, 2015). One editorial linked racial profiling by police to profiling
in other spheres and highlighted the dire consequences of such bias:
For many African Americans, there is fundamentally no difference in the cases.
They see them as part of a fabric of a justice system that doesn’t value Black lives,
in which Driving While Black or Walking While Black or Shopping While Black
leads to harassment which leads to legal problems which leads to unfair arrests
which leads to death. The evidence is on their side. (‘‘In a Nation Holding Its
Breath,’’ 2014, p. A14)
Support for at least part of this statement is found in the Justice Department’s
report on Ferguson’s police department: ‘‘We have found substantial evidence of
racial bias among police and court staff in Ferguson’’ (DOJ, 2015, p. 5). Another
article indirectly touched on racial profiling by discussing the practice whereby
Black parents feel compelled to instruct their children on proper etiquette if they
are stopped by the police:
Many African-Americans consider it a rite of passage to have a talk with their
father or parents about showing respect when stopped by a police officer. Johnson
said he had the conversation with his kids, even though they are not old enough to
drive, knowing that they could be questioned while out walking the dog around the
block. (Bock, 2014, p. A11; cf. Brunson & Weitzer, 2011)
As expected from media reporting on other controversial issues, some of the
assertions might be viewed as hyperbole. For example, Malik Shabazz (President
of Black Lawyers for Justice) was quoted in the Post and Courier as
declaring that ‘‘American police were hunting Black men ‘like a deer or a
dog’’’ (Knapp, 2015d). Such statements illustrate the graphic characterizations
that appeared in some reporting on these incidents.
An additional set of articles (13.7%) discussed race not in terms of racism or
profiling per se, but instead in the context of Racial Disparities in offending
rates, arrest rates, police killings, citizen complaints against officers, or a disjunction
between the city and police department’s racial composition. Regarding
Lee et al. 211
the latter, some articles compared the African American proportion of Ferguson
or North Charleston with the largely White complexion of their police departments.
With respect to complaints lodged against officers, the Post and Courier
reported that the majority of the 120 complaints filed over a recent 4-year period
came from African Americans (Knapp & Smith, 2015). And, regarding racial
disparities in police killings that were not expressly labeled racially biased, one
article recounted a litany of such incidents:
Sadly, we are witnessing incidents of police violence against Black people with
increasing frequency these days. Akai Gurley was killed by police in the stairwell
of his girlfriend’s public housing residence in Brooklyn. We saw Eric Garner
choked to death by a New York City police officer. We witnessed Marlene
Pinnock being pummeled by a California Highway Patrol officer. We saw Levar
Jones shot by a South Carolina police officer during a traffic stop. And just two
days after Michael Brown’s death, Ezell Ford, an unarmed Black man, was killed
by a Los Angeles police officer. (Southerland, 2014, p. A21)
Poverty and Socioeconomic Inequality
Although socioeconomic factors were not as frequently mentioned as the other
items under the Structural Factors rubric, we note that Poverty/Inequality registered
15.4% of mentions across the three cases, and more than one fifth (21.4%)
in Ferguson. Socioeconomic disadvantage was identified as possibly contributing
to racial or class disparities in offending rates, to obtrusive or aggressive
police practices in low-income neighborhoods, or to public distrust of the police.
Consider this ‘‘Why Freddie Gray Ran’’ (2015, p. A22) editorial:
Why did Gray run? He had been arrested a number of times in the past on relatively
minor charges . . . What it makes him is all too typical in a neighborhood
where generations of crushing poverty and the war on drugs combine to rob countless
young people like him of opportunities.’’
And St. Louis Post-Dispatch editor Gilbert Bailon remarked,
What has not grown with [Ferguson’s Black population] is the political representation,
the economic opportunity . . . They feel very isolated and additionally, a
deep mistrust for the police there. Many cities around the country . . . have similar
issues. (Raasch, 2014, p. A4)
The role of socioeconomic inequality is largely absent from the findings of previous
content analyses of media coverage of controversial policing incidents,
suggesting a noteworthy shift in recent coverage of similar events.
212 Police Quarterly 21(2)
The three incidents generated much discussion of reforms to curb officer misconduct
or improve police-community relations. Fully 46.2% of all articles
referred to reforms. Five types were mentioned with some frequency: Body
Cameras, Reform Revenue Intake, Strengthen Police-Community Relations,
Training in Use of Force, and Citizen Review Boards (Table 2).
More than one fifth (21.1%) of articles in the full sample mentioned police
Body Cameras as a type of reform, with significant variation in coverage across
the three incidents: 15.4% St. Louis Post-Dispatch, 51.3% Post and Courier, and
10.4% Baltimore Sun (p<.001). Body cameras were cited as a sensible reform
measure in over half of the articles on Scott’s killing in North Charleston. That
the incident was captured on a citizen’s phone camera, rather than the officer’s
body camera, may explain the Post and Courier’s highlighting the importance of
equipping police officers with body cameras to document future encounters:
Some people fear that, if not for a random video, no one would have been held
accountable for the death of Walter Scott. (Hicks, 2015b)
The video records no instance of Slager telling Scott to stop or halt—the officer just
opened fire and emptied his magazine into a running, unarmed man . . . This is why
police officers should be equipped with body cameras . . . [Rep. Wendell] Gilliard
said last year that, without body cameras, controversies over police shootings will
continue. And this one would have, too – if not for that video. What more proof do
lawmakers need? (Hicks, 2015a)
Despite the Baltimore Sun’s lower proportion of mentions of this reform, the
video footage of Gray being dragged to the police van likely contributed to both
popular criticism of Gray’s treatment and plans to provide officers with body
cameras. Baltimore’s mayor announced that
city officials have assured Gray’s family that police officers in the Western
District—the site of Gray’s arrest—will be the first in the city equipped with
body cameras . . . making sure that something like this doesn’t happen again.
(Wenger & Rector, 2015, p. A1)
All three cities have now begun to implement body-camera policies.
Reform Revenue Intake is a code for content critical of the fines imposed on
offenders and police quotas for arrests or ticketing. Although the former is an
issue with the courts per se, it can also influence police practices, for example, if
an arrest warrant is issued for failure to pay tickets or fines. Some jurisdictions
impose escalating fines or jail time on those who are unable to pay the initial fine.
Revenue reform emerged in a much higher percentage of the articles on
Ferguson (22.5%) than in the other two newspapers because the need for such
Lee et al. 213
reform seemed particularly acute in Ferguson. The local media reported that in
2013 Ferguson’s municipal court ‘‘disposed of 24,532 warrants and 12,018 cases,
or about 3 warrants and 1.5 cases per household’’; fines and court fees comprised
the second largest source of the city’s revenue in 2013, totaling $2,635,400
(‘‘Justice Department investigation,’’ 2014b, p. A12). Another article
reported that ‘‘twenty-five percent of the city’s revenue comes from traffic citations,
prompting a local public defender to argue that most people see the city
as targeting its citizens to raise revenue, not to ‘serve and protect’ them’’
(Thomas, 2014, p. A17).
Ferguson’s practice was branded ‘‘bigotry and profit-driven law enforcement—
essentially using the Black community as a piggy bank to support the
city’s budget through fines’’ (Johnson, 2015, p. A4). In an op-ed in the Post and
Courier, the author noted that this practice occurs in ‘‘many jurisdictions,’’ not
just Ferguson and North Charleston, and concluded, ‘‘We need to stop seeing
the criminal justice system as a source of revenue’’ (Moskos, 2015). The practice
was condemned by the Justice Department in its investigation of Ferguson’s
criminal justice system (DOJ, 2015). Ferguson subsequently lowered the maximum
amount of revenue its municipal courts could contribute to city coffers
from 30% to 10% and cancelled 220,000 of its outstanding arrest warrants for
municipal offenses and traffic violations.
One tenth of the articles commented on the need for more or better Use-of-
Force Training, with significant variation among the three sources (p<.01). It is
not surprising that improved training in use of force and in de-escalating tense
encounters would be a reform discussed in coverage of the incidents, given that
each resulted in a fatality.
The code Strengthen Police-Community Relations refers to the need to
improve mutual understanding and rapport between the police and the population
in general, minority groups, or specific neighborhoods or cities. Some of the
content offered little more than platitudes, such as a statement in the Post and
Courier about the need for more ‘‘outreach efforts’’ by police (Parker, 2015) and
a quotation by North Charleston’s mayor: ‘‘We will be looking for ways to
develop a closer relationship with the individual communities’’ (quoted in
Knapp & Smith, 2015). Other proposals were more specific:
What Ferguson needs are institutions that support cooperation, including a police
force made up of more people who live in the community in order to overcome the
current us vs. them mentality. Officers have to first understand what it means to live
in Ferguson before they can police Ferguson. This kind of cooperation gives people
a stake in their communities. (Thomas, 2014, p. A17)
Ferguson’s mayor promised to increase recruitment of minority officers in order
to ‘‘bridge the gap’’ between the police and the population (Giegerich, 2014,
p. A4), and a newspaper article called for ‘‘mentoring programs at public
214 Police Quarterly 21(2)
schools . . . to steer more minority children into law enforcement careers,’’ which
will create ‘‘positive relationships between the police and the community’’
(McDermott, 2014, p. A3).
The code Citizen Review Board refers to mentions of an agency that has the
authority to review officer behavior after a complaint is filed by a member of the
public, alleging conduct such as excessive force, corruption, harassment, or
verbal abuse. Baltimore and many other big cities have such boards today,
but not Ferguson or North Charleston, whose newspapers presented citizen
review as an imperative reform that would increase police accountability. One
article implied that Ferguson was being too slow to adopt such a board: ‘‘More
than a month after Brown’s death, Ferguson Mayor James Knowles said the city
is still hammering out what its citizen review board will look like’’ (Stuckey,
2014, p. A1). Post and Courier articles discussed the need for a review board for
North Charleston that would have subpoena power to compel officer’s testimony.
Just how vital such boards are perceived to be is reflected in a statement
by James Johnson, President of the North Charleston chapter of the National
Action Network: ‘‘Until we can get a citizens review board, there will never be
any true reform’’ (quoted in Knapp & Elmore, 2015).
This study identified patterns both across the three cities and for each city separately.
Across the cities, as expected, articles discussed the causes of the incident
and who was responsible for the citizen’s death—mirroring reporting on controversial
policing incidents in past decades. But we also found that coverage of
our three incidents included discussions of extra-incident factors: systemic problems,
racial issues, poverty/inequality, and reforms.
For the second research question—differences between the three cases—we
found, first, that some important subcoded content, such as institutional racism,
did not differ statistically between the cases. This may be regarded as a noteworthy
finding, suggesting that institutional racism was recognized as a factor
contributing to police misconduct despite the fact that the circumstances of the
three killings differed. Second, for six of our seven major codes, we found statistically
significant differences between the newspapers. Without speculating
about the reasons for each of these differences, it is clear that some of them
can be explained by features specific to the particular case. The fact that the St.
Louis Post-Dispatch gave attention to the amount of revenue cities raise from
processing minor infractions was clearly rooted in Ferguson’s inordinate use of
its criminal justice system to generate income. Post and Courier articles paid
more attention to reforms, which was largely due to its heavy coverage of
police body cameras—salient because of the visceral power of the video of
Scott’s shooting. The video evidence in Scott’s case also arguably explains
why the Post and Courier was much more likely to attach blame to the officer
Lee et al. 215
than the other two newspapers, where responsibility for the death was less clearcut.
That the Baltimore Sun gave less attention to racial issues may be due to its
police department being much more racially integrated than the other two
departments and to the fact that three of the six officers involved in the
Freddie Gray incident were Black, whereas a single White officer was involved
in the other two incidents.
Prior research suggests that the news media may be more likely to draw
explicit connections between events when they are temporally clustered. We
found some of this cross-fertilization in our data. Nearly 22% of Post and
Courier articles on Walter Scott also mentioned Michael Brown or Freddie
Gray, while 14.6% of Baltimore Sun articles on Gray discussed the Brown or
Scott case. Overall, Brown was the most frequently cross-referenced incident,
garnering 23 total mentions in the Post and Courier (21 mentions alone and 2 in
conjunction with Gray) and 27 mentions in the Baltimore Sun (22 mentions
alone and 5 in conjunction with Scott). All three newspapers also mentioned
incidents in other cities, such as the videotaped killings of Eric Garner in New
York and Tamir Rice in Cleveland. And some articles more generally linked the
current incident to others in the past. The Post and Courier, for example, quoted
a protestor at a demonstration as saying, ‘‘We want the world to understand that
this is not an isolated incident. This has been a reality . . . in the North Charleston
Police Department for many, many years’’ (quoted in Knapp & Smith, 2015). In
contrast to the much more episodic and fragmented coverage of policing incidents
in the past, our three newspapers drew connections between events. Doing
so conveys the message that such incidents are part of a larger problem, beyond
the specific case. The same message may be implicit in the fact that two prominent
mainstream newspapers (Washington Post and The Guardian) began in
2015 to compile publicly available databases with counts and descriptions of
shootings and fatalities at the hands of the police. Whether readers draw connections
between the cases, the fact that these newspapers decided to create these
databases may mean that they, like our three newspapers, no longer normalize
such cases as exceptional.
The news media typically relies on authority figures as primary definers of
events and, in the past structured news regarding police deviance around official
claims. Our findings suggest that this paradigm may be breaking down.
Compared with coverage in past decades, in our three cases newspaper representations
of police killings were more likely to be (a) critical of police practices
and (b) attentive to systemic causes of police misconduct. It is noteworthy that
articles not only attached blame for the incident to the officers involved and
questioned the authorities’ investigation of the incident but also discussed larger
policing practices, race relations, poverty/inequality, and the need for reform. In
other words, it was not unusual for these newspapers to interrogate or draw
damaging conclusions about officer behavior or to highlight structural conditions
that might be responsible for police killings. This critical approach
216 Police Quarterly 21(2)
contrasts sharply with the findings of earlier studies where coverage of similar
types of incidents was skewed in favor of the police and indirectly normalized
police violence. The current study found little pro-police bias or normalization
across our three cases—apart from victim precipitation and victim blaming in
about one seventh of the articles, the exceptional ‘‘rotten apple’’ attribution in
about one twentieth of the articles, and a few other low-prevalence items.
If this shift applies more generally than in our three cases, how can it be
explained? First, it appears that mainstream news reporting has been influenced
by (a) the occurrence of a series of high-profile incidents in a compressed
timespan and (b) the growth of new technologies that the mainstream media
can exploit in their own reporting. The former makes it more difficult to dismiss
an incident as an isolated event or attribute it to a single rogue officer, and
the latter (video recording and social media discourse) can contribute both
evidence and a counter narrative to the police account. Second, newspapers
have a symbiotic relationship with television news reporting. In addition to
national network newscasts, cable news and talk shows provided massive
coverage and commentary on the three incidents as well as the ensuing street
protests in Ferguson and Baltimore. This parallel media coverage may have
had at least some spillover effect on our three mainstream newspapers insofar
as the former highlighted similarities between the incidents and larger policing
problems beyond each specific incident—content that also appeared in our
Our findings suggest that mainstream news reporting may be contributing to
a ‘‘new visibility’’ and critique of police wrongdoing (Goldsmith, 2010). Such
coverage overlaps with other trends, such as growing public criticism of the
police and demands for major reforms. Additional research will help to determine
whether other newspapers, as well as other types of news media, are also
covering the police more critically and systemically than in the past.
Declaration of Conflicting Interests
The authors declared no potential conflicts of interest with respect to the research,
authorship, and/or publication of this article.
The authors received no financial support for the research, authorship, and/or publication
of this article.
1. The proportion of fatal police shootings that were filmed increased from 14% in 2015
to 21% in January to June 2016 (Kindy & Lowery, 2016).
2. An alternative data source would be a national newspaper such as the New York Times
or USA Today. Our selection of local sources for this study was based on our interest
Lee et al. 217
in contextualized coverage of the events in each city (e.g., city-specific explanations for
the incident) as well as the advantage of comparing local content across the three cases.
Another consideration was the likelihood that the local papers would give more coverage
to each case than any national newspaper.
3. The researchers generated the codes based on a systematic and close reading of articles,
and subsequently used Atlas/ti qualitative software to organize the main codes
and distill subcodes. Therefore, the initial coding was manual and grounded and later
coding was computer assisted.
4. Table 2 shows, for example, that 27.7% of articles in the total sample mentioned at
least one of the ‘‘Causes of Incident’’ (see first column). In a similar vein, 15.4% of
articles in the total sample specifically mentioned ‘‘Victim Precipitation,’’ while 5.7%
of articles mentioned ‘‘Racism.’’ If an article mentioned both ‘‘Victim Precipitation’’
and ‘‘Racism,’’ each of these mentions would both be reflected (i.e., coded ‘‘1’’) in the
percentages for each of these respective subcodes. However, because we coded for the
presence or absence of mentions, it would only count once toward the ‘‘Causes of
Incident’’ major code, which is the convention in research allowing for multiple mentions
during the coding process. Table 2 also indicates whether there are statistically
significant differences between the prevalence of mentions of major codes or subcodes
for the three incidents. For example, 44.5% of Post and Courier articles mentioned at
least one of the ‘‘Causes of Incident,’’ compared with 25.5% and 20.3% of articles in
the St. Louis Post-Dispatch and Baltimore Sun, respectively (p<.001).
Ariel, B., Farrar, W., & Sutherland, A. (2014). The effect of police body-worn cameras on
use of force and citizen complaints against the police. Journal of Quantitative
Criminology, 31, 509–535.
Bennett, L. (1990). Toward a theory of press-state relations. Journal of Communication,
Bock, J. (2014, August 31). Mutual respect needed to improve interactions with police,
panelists say. St. Louis Post-Dispatch, p. A11.
Brown, G. (2016). The blue line on thin ice: Police use of force modifications in the era of
camera phones and YouTube. British Journal of Criminology, 56, 293–312.
Brunson, R.K. & Weitzer, R. (2011). Negotiating unwelcome police encounters: The
intergenerational transmission of conduct norms. Journal of Contemporary
Ethnography, 40, 425–456.
Chermak, S., & Weiss, A. (2005). Maintaining legitimacy using external communication
strategies: An analysis of police-media relations. Journal of Criminal Justice, 33,
Chermak, S., McGarrell, E., & Gruenewald, J. (2006). Media coverage of police misconduct
and attitudes toward police. Policing, 29, 261–281.
Cohn, M. (2015, May 1). Several factors in medical evidence; Experts say autopsy, video,
more will help determine how spine was injured. Baltimore Sun, p. A14.
Condon, S. (2015, August 11). A year after Ferguson, Washington still working on police
reforms. CBS News. Retrieved from http://www.cbsnews.com/news/a-year-after-ferguson-
Cook, T. (1998). Governing with the news. Chicago, IL: University of Chicago Press.
218 Police Quarterly 21(2)
Darby, J. (2015, June 11). Get real about police bias and White privilege. Post and
Courier. Retrieved from http://www.postandcourier.com/opinion/get-real-aboutbiased-
Deere, S. (2014, November 19). Nixon announces members of Ferguson Commission. St.
Louis Post-Dispatch, p. A1.
Department of Justice. (2015). Investigation of the Ferguson Police Department. U.S.
Department of Justice, Civil Rights Division. Retrieved from https://www.justice.
Department of Justice. (2016). Investigation of the Baltimore City Police Department. U.S.
Department of Justice, Civil Rights Division. Retrieved from https://www.justice.gov/
Elmore, C., & MacDougall, D. (2015, April 4). Man shot and killed by North Charleston
police officer after traffic stop; SLED investigating. Post and Courier. Retrieved from
Ericson, R. (1995). The news media and account ability in criminal justice. In P.
C. Stenning (Ed.), Accountability for criminal justice (pp. 135–161). Toronto, ON:
University of Toronto Press.
Eschholz, S., Blackwell, B., Gertz, M., & Chiricos, T. (2002). Race and attitudes toward
the police: Assessing the effects of watching ‘‘reality’’ police programs. Journal of
Criminal Justice, 30, 327–241.
Giegerich, S. (2014, December 1). Ferguson mayor announces plan to increase minority
hiring and no severance for Wilson. St Louis Post-Dispatch, p. A4.
Goldsmith, A. (2010). Policing’s new visibility. British Journal of Criminology, 50,
Goold, B. (2003). Public area surveillance and police work: The impact of CCTV on
police behavior and autonomy. Surveillance and Society, 1, 191–203.
Hicks, B. (2015a, April 8). Slager’s actions not only devastate a family, but an entire
city. Post and Courier. Retrieved from http://www.postandcourier.com/staff/brian_
Hicks, B. (2015b, April 9). Walter Scott video should change everything. Post and
Courier. Retrieved from http://www.postandcourier.com/staff/brian_hicks/thisshould-
Hirschfield, P., & Simon, D. (2010). Legitimizing police violence: Newspaper narratives
of deadly force. Theoretical Criminology, 14, 155–182.
Improved dialogue on police. (2015, May 5). Post and Courier. Retrieved from http://
In a nation holding its breath, Ferguson is everywhere. (2014a, December 5). St. Louis
Post-Dispatch, p. A14.
Iyengar, S., & Kinder, D. (1987). News that matters. Chicago, IL: University of Chicago Press.
Jennings, W., Lynch, M., & Fridell, L. (2015). Evaluating the impact of police officer
body-worn cameras on response-to-resistance and serious external complaints. Journal
of Criminal Justice, 43, 480–486.
Lee et al. 219
Johnson, G. (2015, March 9). Justice Department’s Ferguson report resonates across the
US. St. Louis Post-Dispatch, p. A4.
Justice Department investigation will have to expand beyond Ferguson. (2014b,
September 5). St. Louis Post-Dispatch, p. A12.
Kaminski, R., & Jefferis, E. (1998). The effect of a violent televised arrest on public
perceptions of the police. Policing, 21, 683–706.
Kindy, K., & Lowery, W. (2016, July 8). Deadly shootings by police on the rise in 2016.
Washington Post, pp. A1, A8–A9.
Knapp, A. (2015a, August 24). Michael Slager, officer charged in Walter Scott’s death,
likely to wait longer for bond hearing. Post and Courier. Retrieved from http://www.
Knapp, A. (2015b, July 13). National NAACP group leaders join call for federal review
of North Charleston police practices. Post and Courier.Retrieved from http://www.
Knapp, A. (2015c, April 7). North Charleston officer faces murder charge after video
shows him shooting man in back. Post and Courier. Retrieved from http://www.postandcourier.
Knapp, A. (2015d, April 12). Sharpton: In wake of Walter Scott’s death, North
Charleston could set ‘‘new tone in policing’’ nationwide. Post and Courier.
Retrieved from http://www.postandcourier.com/archives/rev-sharpton-praises-cityon-
Knapp, A., & Elmore, C. (2015, May 2). Call for police oversight board persists while
North Charleston resists. Post and Courier. Retrieved from http://www.postandcourier.
Knapp, A., & Smith, G. (2015, April 8). Shooting death of Walter Scott deals blow to
North Charleston’s struggle to balance civil rights, public safety. Post and Courier.
Retrieved from http://www.postandcourier.com/archives/shooting-death-of-walterscott-
Kohler, J. (2014, August 14). Key witness meets with authorities to discuss Brown shooting.
St. Louis Post-Dispatch, p. A9.
Lawrence, R. (2000). The politics of force: Media and the construction of police brutality.
Berkeley: University of California Press.
Marbella, J. (2015, May 2). States Attorney stuns public with prompt charges: Six
charged. Baltimore Sun, p. A1.
McCombs, M., & Shaw, D. (1972). The agenda-setting function of the mass media. Public
Opinion Quarterly, 36, 176–187.
McDermott, K. (2014, October 2). Post-Ferguson panel suggests more police mentoring
of minority students. St. Louis Post-Dispatch, p. A3.
Mosby’s ‘‘conflicts.’’ Our view: The prosecutor’s conduct hasn’t been flawless, but suggestions
that she must recuse herself from the Freddie Gray case are baseless. (2015,
May 21). Baltimore Sun, p. A16.
220 Police Quarterly 21(2)
Moskos, P. (2015, April 14). Scott’s death was not another Ferguson. Post and Courier.
Retrieved from http://www.postandcourier.com/opinion/scott-s-death-was-notanother-
Paletz, D., & Entman, R. (1981). Media power politics. New York, NY: Free Press.
Parker, A. (2015, April 20). The cops and the kids: Can community outreach transform
policing? Post and Courier. Retrieved from http://www.postandcourier.com/archives/
Partisan melee over law enforcement is a disgrace. (2015, January 30). St. Louis Post-
Dispatch, p. A14.
Pew Research Center. (2016). The modern news consumer. Washington, DC: Pew.
Pew Research Center. (2017). Behind the badge. Poll of police officers, May 19-August 14,
2016, N¼7,917; poll of the public, August 16-September 12, 2016, N¼4,538. Retrieved
Pollack, E., & Allern, S. (2014). Criticism of the police in the news. Nordicom Review, 35,
President’s Task Force on 21st Century Policing. (2015). Final report. Washington, DC:
Office of Community Oriented Policing Services.
Raasch, C. (2014, August 18). Nixon calls on prosecutors to ‘‘step up’’ investigation.
St. Louis Post-Dispatch, p. A4.
Sigelman, L., Welch, S., Bledsoe, T., & Combs, M. (1997). Police brutality and public
perceptions of racial discrimination: A tale of two beatings. Political Research
Quarterly, 50, 777–791.
Simonson, J. (2016). Copwatching. California Law Review, 104, 391–445.
Southerland, V. (2014, November 26). Addressing the scourge of police violence. St.
Louis Post-Dispatch, p. A21.
Strauss, A., & Corbin, J. (1990). Basics of qualitative research. London, England: Sage.
Stuckey, A. (2014, September 22). Ferguson, St. Louis contemplates police review board.
St. Louis Post-Dispatch, p. A1.
Thomas, M. (2014, August 22). Building community in Ferguson. St. Louis Post-
Dispatch, p. A17.
Victor, D., & McPhate, M. (2016, July 8). Critics of law enforcement embrace Facebook
Live and other video tools. New York Times, p. A18.
Waegel, W. (1984). How police justify the use of deadly force. Social Problems, 32,
Washington Post/ABC News. (2014). Poll, December 11-14. N¼1,012. Retrieved from
Weisburd, D., Greenspan, R., Hamilton, E., Bryant, K., & Williams, H. (2001). The abuse
of police authority: A national study of police officers’ attitudes. Washington, DC:
Weitzer, R. (2015). American policing under fire: Misconduct and reform. Society, 52,
Weitzer, R., & Tuch, S. (2006). Race and policing in America: Conflict and reform.
New York, NY: Cambridge University Press.
Lee et al. 221
Wenger, Y., & Rector, K. (2015, September 10). Hearings to resume. Baltimore Sun,
What took so long? Our view: The need for more than the Department of Justice’s
‘‘collaborative review’’ of the Baltimore Police Department was evident long before
Freddie Gray’s death. (2015, May 7). Baltimore Sun, p. A20.
Why Freddie Gray ran. Our view: We don’t yet know what caused Freddie Gray’s
injuries, but we know all too well the forces that brought him together with six
police officers that day. (2015, April 26). Baltimore Sun, p. A22.
Wilson, R. (2015, February 8). Police accountability measures flood state legislatures
after violent events. Washington Post. Retrieved from https://www.washingtonpost.
Wooten, F. (2015, April 14). No hiding from vile video truth. Post and Courier. Retrieved
Angela S. Lee received her MA in Criminology at George Washington
University in 2016. She received the Outstanding Graduate Student award
from the sociology department in 2016, and she plans to pursue her doctoral
education in the near future.
Ronald Weitzer is a professor of sociology at George Washington University. He
has published extensively in the area of policing, based on research in the United
States, South Africa, Israel, and Northern Ireland. His books include Race and
Policing in America: Conflict and Reform and Policing Under Fire: Ethnic
Conflict and Police-Community Relations in Northern Ireland.
Daniel E. Martı´nez is an assistant professor of sociology at the University of
Arizona. He received his PhD from the University of Arizona, and his research
interests include unauthorized migration, human smuggling, youth gangs, and
Daniel E. Martınez, Angela S. Lee, Ronald Weitzer, Police Quarterly 2018, Vol. 21(2) 196–222, “Recent Police Killings in the United States: A Three-City Comparison”, https://www.researchgate.net/profile/Ronald_Weitzer/publication/322007819_Recent_Police_Killings_in_the_United_States_A_Three-City_Comparison_2018/links/5ae9870e0f7e9b837d3bab05/Recent-Police-Killings-in-the-United-States-A-Three-City-Comparison-2018.pdf?origin=publication_detail