American Police Brutality in Black and White: Fox News Website Coverage of the Killings of George Floyd and Tyre Nichols

Abstract

The police-perpetrated killings of George Floyd in 2020 and Tyre Nichols in 2023, respectively, provide a unique opportunity for comparison of news content. Both incidents were recorded on video and appear to be characterized by unwarranted aggression and brutality. A key difference worth exploring is the race of the perpetrator(s). In the case of Floyd, the killer, Derek Chauvin, was white, while Nichols’ killers were black. The purpose of the present research was to comparatively examine how the Fox News website framed these two police-perpetrated killings. Quantitative content analysis was used to examine Fox News website coverage of the murders of Floyd and Nichols during the one-month periods following the publicity of both incidents. Findings were consistent with researcher expectations that Fox News would be more sympathetic to Nichols than Floyd, and more critical of the black police officers who killed Nichols than the white officer who killed Chauvin. In all, eight out of the study’s ten sub-hypotheses were supported.

Keywords:

• African-Americans
• Framing
• Content Analysis
• George Floyd
• Black Lives Matter

Introduction

The May 25, 2020 murder of George Floyd, a 46-year old African-American, by white Minneapolis police officer Derek Chauvin sparked global protests and national and international discussions about American police brutality and racism (Roth & Jarrar, 2021). Public anger became widespread when a video of Chauvin kneeling on Floyd’s neck for more than 9-min, until his death, was made public.

Less than three years later, on January 7, 2023, Tyre Nichols, a 29-year-old African American, was killed by five black police officers in Memphis, Tennessee (Sainz & Mattise, 2023). Although Nichols’ death at the hands of police did not cause the same degree of public uproar that Floyd’s murder did, it provoked interesting and important discussions about broader patterns of police brutality, the role of race in American policing, police training, and other matters (Cranmore, 2023).

Previous scholarship has scrutinized the content of Fox News, America’s most viewed cable news network, as apparently lacking in professionalism (Aday 2010; Harmon & Muenchen, 2009; Peck, 2014; Peters, 2010), and generally supporting anti-black narratives (Hodges, 2015; el-Nawawy & Elmasry, 2018; Mills, 2017; Shah & Yamagami, 2015; Zhang, 2021), among other criticisms.

The killings of Floyd and Nichols, respectively, provide a unique opportunity for comparison of Fox News content. Both acts were carried out by police, were recorded on video, and appear to be characterized by unwarranted aggression and brutality. A key difference worth exploring is the race of the perpetrator(s). In the case of Floyd, the killer, Chauvin, was white, while Nichols’ killers were black.

The purpose of this research is to comparatively examine how the Fox News website framed these two police-perpetrated killings. How did the Fox News website frame each killing? To what extent did Fox News’ apparent underlying sympathy with white supremacy color its coverage? What sorts of overlaps and tensions exist in the coverage patterns of the two events? What explains the overlaps and tensions? These are the broad questions which the present study seeks to answer.

In particular, this paper examines Fox News website coverage of the murders of Floyd and Nichols during the one-month periods following the publicity of both events. Framing theory is used to guide the study, which employs quantitative content analysis to test several hypotheses related to victim and perpetrator framing, sourcing, and patterns of legitimation and condemnation.

Literature review

U.S. Media and race

Media framing of race plays an integral role in how public opinion perceives and socially constructs race (Kaufmann, 2021). This is especially the case in the context of U.S. media, which tend to produce and (re)produce frames that strengthen prevailing ideological trends (Chen et al., 2022).

Racism has been a prominent feature in a majority of studies examining U.S. media coverage of race. Racism can be defined as “a complex societal system in which peoples of European origin dominate peoples of other origins…” (van Dijk, 1999, p. 24). Racism often leads to a bipolar environment, where members of the prevailing white “ingroup” formulate racist predispositions and negative stereotypes about members of the less dominant, nonwhite “outgroup” (van Dijk, 2000, p. 44).

Scholarly literature points to the manifestation of this bipolar environment in U.S. media framing of race issues. For example, a qualitative analysis of mainstream American media coverage of the aftermath of Hurricane Katrina that hit the southeast part of the United States in 2005 revealed an inclination to connect African Americans with criminal behavior, violence and looting that took place after the hurricane. This reinforced negative stereotypical perceptions about Blacks in America (Sommers et al., 2006).

Another study examined New York Times and Washington Post coverage of the fatal 2012 shooting of Trayvon Martin, an African American teenager, at the hands of George Zimmerman, a neighborhood volunteer watchman. Findings showed that coverage “reflected an encompassing pro-white/anti-black master-frame that presented Black Americans as inadequate, lawless, criminal, threatening and at times biologically different…Conversely, whites were presented favorably as ‘protectors’ and ‘virtuous’” (Lane et al., 2020).

Some research has examined how U.S. media have covered Black Lives Matter (BLM), a movement formed in 2013 to protest perceived systematic injustices against African Americans (Umamaheswar, 2020). Most studies show evidence of negative framing of BLM.

For instance, a qualitative textual analysis of American newspapers revealed that although the movement’s objectives were presented in a positive manner, the framing of the movement was both dramatized and politicized. Moreover, newspapers focused more on the movement’s negative rather than positive outcomes (Umamaheswar, 2020).

Reid and Craig (2021) content analysis found that most articles in six American newspapers highlighted the allegedly chaotic and disorderly nature of BLM protests. The newspapers under study also framed BLM protesters as a threat to public safety (Reid & Craig, 2021).

A visual analysis of 200 images gathered from the websites of three mainstream U.S. broadcast networks, CNN, MSNBC and Fox News, found that the images focused disproportionately on violence (Gondwe & Bhowmik, 2022, p. 1).

Not all studies dealing with U.S. media coverage of the BLM movement found evidence of negative framing. A content analysis of New York Times and St. Louis Post-Dispatch coverage of the 2014 police killing of Michael Brown in Ferguson, Missouri found that both papers privileged protester sources over police sources and adopted an overall positive framing of the movement (Elmasry & el-Nawawy, 2017).

U.S. media coverage of police brutality

Incidents of police brutality against civilians in the United States have increased in recent years. The year 2022 alone witnessed 1,097 fatal police shootings. The rate of such shootings was proportionately higher among Black Americans than other racial and ethnic groups (Statista, 2023). Despite the recent spike in the number of fatal shootings by police, few studies have analyzed U.S. media framing of police brutality. The handful of studies that have been conducted seem to point to pro-white racial biases in media coverage.

A content analysis was carried out on the New York Times and Washington Post framing of the 1991 police beating of Rodney King, an African American male, at the hands of police in Los Angeles. While coverage in both papers tended to point to problems in the Los Angeles Police Department, it overlooked the systemic roots of anti-black racism in American society and portrayed King negatively and stereotypically (Solomon, 2004, p. 23).

A content analysis of 105 articles published in American newspapers between 1997 and 2000 revealed that victims of police brutality were framed as “social threats.” Moreover, sampled articles “situate[d] police actions within legitimate institutional roles” (Hirschfield & Simon, 2010, p. 155).

Another study (Mourão et al., 2021) looked into U.S. media coverage of the protests that followed the aforementioned 2014 killing of Brown in Ferguson. Findings suggest that while initial coverage adopted a violence frame aimed at allowing police to use force, coverage patterns shifted over time toward a focus on highlighting protestor voices and demands.

A content analysis of Washington Post news articles dealing with police violence against Black women and men pointed to gender differences in coverage. While excessive police violence was blamed for many cases of black male suffering, black women were undermined. In most examined articles, in fact, black female victimization was dismissed outright (Hollingshead, 2022, p. 346).

A study of New York Post and New York Daily News coverage of police brutality in New York pointed to explicit biases against the Black Lives Matter movement. The New York Post linked Black Americans with felonious behavior and tended to justify police brutality (Chama, 2019).

Background on the George Floyd and Tyre Nichols killings

Following a complaint to the Minneapolis Police Department on May 25, 2020 about a counterfeit bill at a local business, four white police officers were sent to the business location in Minneapolis, Minnesota, where a Black American male, George Floyd, was sitting in a car parked by the business. After complying with police orders to exit the car, Floyd was handcuffed and taken to a squad car. He struggled, however, citing claustrophobic apprehension (Dixon & Dundes, 2020).

Inside the police car, Floyd suffered from breathing problems. He was then taken out and placed on the sidewalk, where he was surrounded by all four officers. One of the officers, Derek Chauvin, who is white, placed his knee on Floyd’s neck. Floyd repeatedly expressed his inability to breathe and pleaded for help, but Chauvin continued to press his knee on Floyd’s neck. This position was maintained for more than eight minutes. By the time Chauvin lifted his knee, Floyd was unconscious. Paramedics arrived at the scene, but Floyd was pronounced dead as a result of cessation of heart and respiration functions. Video of Floyd lying under Chauvin and crying for help went viral on social media, leading to wide-scale protests locally and nationally (Dixon & Dundes, 2020). All four officers were convicted of federal and state charges (Parks & Levenson, 2023), with Chauvin ultimately convicted of murder and sentenced to more than 20 years in prison (Karnowski, 2023).

On January 7, 2023, Tyre Nichols, a 29-year-old African American man, was stopped for reckless driving by five Black police officers in Memphis, Tennessee. The officers started violently kicking and beating Nichols, who was then transferred to the hospital in critical condition. Nichols passed away three days after the incident. Video of the graphic incident was made public shortly after the incident, causing public anger. The five officers were fired from their jobs and pleaded not guilty to several felony charges filed against them (Rojas et al., 2023).

Fox News

Created in 1996 by Australian media entrepreneur Rupert Murdoch, Fox News has become a key asset of the international media network Fox Corporation (Ray, 2024). Fox News is among the most prominent, influential and widely followed 24-h cable television networks in the United States. In fact, it has enjoyed the highest ratings of all American cable news channels over the past several years, with more than four million viewers tuning in regularly to nightly prime-time programming (Ash et al., 2022). It is a right-wing network with a conservative ideology that caters to the Republican socio-political agenda. A recent Pew poll showed that approximately 65% “of Republican and Republican-leaning voters say they trust Fox News for political and election news” (Bauer et al., 2022, p. 20).

Despite accusations by Fox News’ critics that its coverage and programming reflect partisan bias, the network markets itself as “fair and balanced” (Bard, 2017, p. 102). The disconnect between the network’s self-proclaimed description and its critics’ questioning of its journalistic integrity contributes to the importance of analyzing its content (Bard, 2017).

When it comes to its coverage of race-related issues, a series of studies have found that Fox News has covered race “through a prism of underlying ideologies, specifically those of white supremacy and racial capitalism” (Mills, 2017, p. 40). In this context, a qualitative textual analysis of how Fox’s Hannity Show framed the ‘Unite the Right’ rally that took place in Charlottesville, Virginia in 2017 showed that the program defended President Donald Trump, sided with Republican arguments, and cast blame upon Black people and American liberals (el-Nawawy & Elmasry, 2023).

A qualitative study of Fox News programming looked into the network’s coverage of America’s 2011 budget crisis. The analysis, which covered 36 h of content, found that Fox hosts subtly undermined the integrity of minority correspondents, especially Black correspondents, who reported on the issue (Moore, 2011, p. 64).

An analysis of Fox News coverage of the protests that followed the 2014 police killing of Brown in Ferguson showed that Fox systematically belittled and criminalized Black protestors by using such keywords as “thugs,” “looting,” “mayhem” and “chaos” (Mills, 2017, p. 50). The analysis also showed that the network went out of its way to accuse Black Americans of creating racial disparities in the society (Mills, 2017).

A content analysis of the coverage of the protests associated with the BLM movement on four American cable networks, including Fox News, found that Fox focused the most on the protestors’ alleged disorder, unruliness, and lack of restraint. Finding also showed that Fox rarely covered any violent acts that were committed by police against protestors (Brown & Mourão, 2022).

A study analyzing discourse on the BLM movement in both CNN and Fox News found that CNN was more favorable toward the movement than Fox News, which had a higher tendency to associate the movement with Marxist and terrorist ideologies. In most of its references to the movement, Fox News used terms such as “rioters, riots, militants, screamers, terrorists, thugs, violence, and zealots” (Repo, 2022, p. 76).

Some critics have noted that Fox is known for its dismissal, or at least skepticism, of the argument that there is racial inequality or racism in American society (Lapierre & Aubrey, 2022; Moore, 2011). In perpetuating this narrative, Fox News is said to hamper “meaningful discussion of racism in the [U.S.] criminal justice system” (Mills, 2017, p. 52).

Framing theory

The origin of framing can be attributed to sociologist Erving Goffman (1974), who defined it as a “schemata of interpretation” that helps individuals understand or perceive an issue or a phenomenon (p. 21). Building on Goffman’s definition, Entman (1993) argued that framing entails the selection of “some aspects of a perceived reality and mak[ing] them more salient in a communicating text, in such a way as to promote a particular problem definition, causal interpretation, moral evaluation, and/or treatment recommendation” (p. 52).

Framing is an integral component of the media field, as news frames contribute to “turn[ing] meaningless and nonrecognizable happenings into a discernible event” (Scheufele, 1999, p. 106). Framing refers to a dynamic, rather than a static, process, in the sense that there are ongoing interactions between news frames and audiences’ predilections (de Vreese, 2005).

Journalists often subconsciously adopt specific news frames as they package, edit and present stories, a reality which may produce effects in audience perceptions (Price et al., 1997). During this process, several devices are utilized to exemplify and drive news frames. Included among these are photographs, sources, images, headlines, tone, phrases, illustrations and metaphors, all of which can help anchor frames (Pan & Kosicki, 1993). Audiences are affected by news frames without identifying the tactics that were used in structuring them (Entman, 1993).

One key framing aspect is the tone-of-voice in a news story, which tends to affect how the story’s content is perceived and interpreted (Kuttschreuter et al., 2011). Embedded in a given news story’s tone are the biases, inclinations and cultural nuances of the reporter(s) who worked on it (Brunken, 2006).

A story’s overall tone is directly tied to the type of sources that are used in it. Sources play an integral role in how news stories are framed, and can lead to either magnifying or marginalizing positions, issues, and groups (Gonçalves, 2023). Regarding sources, Strömbäck et al. (2013) said, “Who gets to speak in the news ultimately decides what audiences may hear from the news. This is why the question of news sources is also a question of power and influence and why it is important to investigate the impact of the media’s use of sources” (p. 33–34). The credibility levels and the social class of news story sources affect overall perceptions of news frames (Coleman et al., 2011). A news story’s tone and sourcing can have an impact on the overall directionality of a given frame. This directionality can produce both attitudinal and behavioral effects in news audiences (Reiter-Haas, 2023).

Hypotheses and research questions

Based on framing theory and past scholarly literature on Fox News, the following hypotheses and research questions are put forth:

1. H1. The Fox News website will be more sympathetic to Tyre Nichols than George Floyd.

   1. H1a. Fox News articles will be more likely to frame Tyre Nichols as a victim than George Floyd.

   2. H1b. Fox News articles will be more likely to include quotes and paraphrases from the family and friends of Tyre Nichols than the family and friends of George Floyd.

   3. H1c. Fox News articles will provide more personal details about Tyre Nichols than George Floyd, on average.

2. H2. The Fox News website will be more sympathetic to police officers during its coverage of the George Floyd incident than in its coverage of the Tyre Nichols incident.

   1. H2a. Fox News articles will be more likely to frame police officers as honorable crime fighters during coverage of the George Floyd incident than in coverage of the Tyre Nichols incident.

   2. H2b. Fox News articles will be more likely to include quotes and paraphrases from police officers during coverage of the George Floyd incident than in coverage of the Tyre Nichols incident.

   3. H2c. Fox News articles about the George Floyd incident will be less likely to include expert condemnations of the police force than articles about the Tyre Nichols incident.

   4. H2d. Fox News articles will be less likely to describe the George Floyd incident as “police brutality” than the Tyre Nichols incident.

   5. H2e. Fox News articles will provide more personal details about police during coverage of the George Floyd incident than during coverage of the Tyre Nichols incident, on average.

   6. H2f. Fox News articles will be less likely to suggest that police officers overstepped against George Floyd than Tyre Nichols.

   7. H2g. Fox News articles will be more likely to suggest that police violence was justified against George Floyd than Tyre Nichols.

RQ1: How does Fox News talk about anti-police protests during coverage of both incidents? Will a more critical tone be adopted against anti-police protests following the Floyd incident?

RQ2: What is the overall impression provided by Fox News about police in general during coverage of both incidents? Will the website’s coverage be more sympathetic toward police in the context of the Floyd incident?

Method

To test the hypotheses and answer the research questions, this study employed quantitative content analysis, a method which is ideal for comparing relatively large quantities of news (and other) content. Quantitative content analysis involves using rules of measurement to assign numerical value to content patterns and categories, and the employment of statistical tools to examine data (Riffe et al., 2005).

Since the basic purpose of the study was to examine Fox News website coverage of two separate events—the May 2020 killing of George Floyd and the January 2023 killing of Tyre Nichols, respectively—Fox News was the lone news site selected for analysis. The unit of analysis was the individual article.

The one-month periods following the publicity of both killings were selected as sampling time periods. For the Floyd incident, this corresponded to May 25, 2020, the day of Floyd’s death, through June 24, 2020. For the Tyre Nichols event, the dates were January 16, 2023, nine days after Nichols was killed, through February 15, 2023. [The reason for the gap between the date of Nichols’ death and the start of the sampling time period is that news of Nichols’ killing was not widely publicized in the immediate aftermath of the incident. Rather, the killing did not become a national story until January 16, 2023.]

Initially, researchers sought to use the Fox News website to retrieve articles. However, the website’s search engine does not include a comprehensive archive of past articles. The researchers ultimately used the Google archive as a source for articles.

The search terms “George Floyd Fox News” and “Tyre Nichols Fox News” were used to retrieve articles using the Google search engine. For each one-month period, all Fox News articles dealing centrally with the topic(s) were selected for analysis. Both editorials and news articles were sought. The final sample consisted of 223 articles, 128 about the Floyd killing and 95 about the Nichols killing. Nearly all examined articles (92.4%) were news articles, with very few editorials (7.6% of the total number of articles) published about either event.

Coding

The coding scheme included items assessing how the article framed the victim; how the article framed police; the extent to which the article humanized the victim; the extent to which the article humanized police officers; the overall directionality of the article toward the victim; overall directionality of the article toward police; and the number of times the article sourced police officers and the victim’s family and friends, respectively. Additionally, some coding sheet items asked about race, racism, and police brutality.

The framing the victim question was measured at the nominal level and asked coders to consider the primary emphasis of the article. Articles that focused primarily on the victim’s alleged criminal activity were coded “as a criminal,” while articles that focused primarily on either the brutality of the police officers or resulting injury were coded “as a victim.”

The framing police question focused on textual descriptions of police officers and their actions. Coders were instructed to code “as criminals/aggressors” any articles that focused primarily on apparent police transgressions and “as honorable crime fighters” any articles that emphasized the heroism of the officers, the violence of the black victim, and/or apparent dangers to police.

The humanization questions asked coders to tally all mentions of name, age, hobbies, family relations, and occupation/profession.

Directionality questions were concerned with total, comprehensive readings of the articles and whether readers would be more likely to come away from an individual article with a positive or negative impression of the subject. Articles that focused primarily on positive subject aspects (good traits, helpful or kind actions, or vulnerability) were coded sympathetically, as “positive.” Articles that focused primarily on negative subject aspects (bad traits, poor behavior, or aggression) were coded unsympathetically, as “negative.” If articles focused equally on negative and positive subject aspects, coders were instructed to consider the direction of the first piece of information (e.g. information found in the headline or lead).

Coding sheet items about sourcing required coders to count up the total number of times subjects were sourced. For the purposes of these questions, both paraphrases and direct quotes were counted as instances of sourcing.

Two media studies graduate students served as coders. Training on the coding scheme was carried out in late Spring 2023. Intercoder reliability testing and final coding were carried out at the immediate conclusion of training. All coding was completed by early summer 2023.

Variables were measured at either the nominal or ratio levels. Intercoder reliability testing was carried out on 20% of the final sample. Reliability scores were calculated using the ReCal online tool (Freelon, 2010). Scott’s Pi was calculated for all nominal level variables, while Krippendorff’s Alpha was calculated for each ratio level variable. Perfect reliability was achieved for all variables except three: a variable about the overall directionality of the article toward the victim which was measured at the nominal level; a nominal level variable assessing whether the article justified the police stop; and a ratio level variable measuring how many times the incident was described as an act of “police brutality.” Reliability for these three variables ranged from good to excellent. The directionality toward the victim variable realized a 0.73 Scott’s Pi score; the justification of the police stop variable achieved a 0.93 Scott’s Pi score; and the “police brutality” mentions variable accrued a 0.97 Krippendorff’s Alpha score.

Findings

Findings generally show support for the hypotheses and sub-hypotheses. There was partial support for H1, which predicted more sympathetic coverage for Tyre Nichols than George Floyd. In all, two of H1’s three sub-hypotheses were supported. There was much more support for H2, which predicted more sympathetic coverage for police officers in the context of the Floyd incident than the Nichols incident. For H2, six out of seven sub-hypotheses were supported.

H1 predicted that “the Fox News website would be more sympathetic to Tyre Nichols than George Floyd.” H1 included three sub-hypotheses, two of which were supported.

H1a predicted that Fox News articles would be more likely to frame Nichols as a victim than Floyd. This sub-hypotheses was not supported. Articles overwhelmingly framed both Floyd (98.4%) and Nichols (100%) as victims and differences were not statistically significant [χ² (df = 1, N = 223) = 1.5, p = .221]. Table 1 displays the results.

Table 1. Framing of victim.

	Dominant Frame of Victim
Event	As a Victim	As a Criminal	Total
George Floyd	126 (98.4%)	2 (1.6%)	128 (100%)
Tyre Nichols	95 (100.0%)	0 (0.0%)	95 (100%)
Total	221 (99.1%)	2 (0.9%)	223 (100%)

χ² (df = 1, N = 223) = 1.5, p = .221.

H1b predicted that articles would be more likely to include quotes and paraphrases from the family and friends of Nichols than from the family and friends of Floyd. This sub-hypothesis was supported. Articles about the Nichols incident quoted his family and friends an average of 2.04 times per article. Meanwhile, articles about the Floyd incident quoted his family and friends 1.17 times per article, on average. As shown in Table 2, this difference was statistically significant [t(223) = −1.64, p = .05].

Table 2. Mean number of quotes and paraphrases from family/friends of the victim.

Event	Mean quotes and paraphrases from family/friends of victim
George Floyd	1.17 (SD = 3.6)
Tyre Nichols	2.04 (SD = 4.3)

Mean number of quotes/paraphrases from family/friends of victim: t(df = 221, N = 223) = 1.64, p = .05.

H1c predicted that Fox News articles would provide more personal details about Nichols than Floyd, on average. This sub-hypothesis was also supported. Articles about the Nichols killing provided more personal details about him (M = 3.31) than articles about the Floyd killing (M = 1.90). A t-test showed the difference to be statistically significant at the 0.05 level [t(223) = 2.39, p = .009]. Results are displayed in Table 3.

Table 3. Mean number of personal details about the victim.

Event	Mean Number of Personal Details about Victim
George Floyd	1.90 (SD = 1.2)
Tyre Nichols	3.31 (SD = 5.6)

Mean number of personal details about victim: t(df = 100.5, N = 223) = 2.39, p = .009.

H2 predicted that the Fox News website would be more sympathetic to police officers during its coverage of the George Floyd incident than during its coverage of the Tyre Nichols incident. H2 included seven sub-hypotheses, six of which were supported.

H2a predicted that Fox News articles would be more likely to frame police officers as honorable crime fighters during coverage of the Floyd incident than in coverage of the Nichols incident. This sub-hypothesis was supported. A total of 14.1% (N = 18) of articles covering the Floyd incident framed police as honorable crime fighters, with the remaining articles (85.9%) framing police as criminals/aggressors. None of the articles covering the Nichols incident framed police as honorable crime fighters. Results are displayed in Table 4.

Table 4. Framing of police.

	Dominant frame of police
Event	As Criminals/Aggressors	As honorable crime fighters	Total
George Floyd	110 (85.9%)	18 (14.1%)	128 (100%)
Tyre Nichols	95 (100.0%)	0 (0.0%)	95 (100%)
Total	205 (91.9%)	18 (8.1%)	223 (100%)

χ² (df = 1, N = 223) = 14.5, p < .001.

H2b predicted that Fox News articles would be more likely to include quotes and paraphrases from police officers during coverage of the Floyd incident than during coverage of the Nichols incident. This sub-hypothesis was supported. Results, which are displayed in Table 5, show that police were quoted more in the context of coverage of Floyd (M = 2.49) than in the context of coverage of Nichols (M = 2.04). A t-test showed that this difference was statistically significant at the .05 level [t(223) = 1.85, p = .03].

Table 5. Mean number of quotes and paraphrases from police.

Event	Mean quotes and paraphrases from police
George Floyd	2.49 (SD = 6.9)
Tyre Nichols	2.04 (SD = 4.3)

Mean number of quotes/paraphrases from police: t(df = 221, N = 223) = 1.85, p = .03.

H2c predicted that Fox News articles about the Floyd incident would be les likely to include expert condemnations of the police force than articles about the Nichols incident. This sub-hypothesis was also supported. Only 11% of articles (N = 14) about the Floyd incident included expert condemnations of force, compared to 25.3% of articles (N = 24) about the Nichols incident. Table 6 displays the results of the differences, including a chi-square test showing statistical significance [χ² (df = 1, N = 222) = 7.8, p = .005].

Table 6. Expert condemnation of use of force.

	Expert condemnation of use of force
Event	Expert condemned use of force	Expert did not condemn use of force	Total
George Floyd	14 (11.0%)	113 (89.0%)	127 (100%)
Tyre Nichols	24 (25.3%)	71 (74.7%)	95 (100%)
Total	38 (17.1%)	184 (82.9%)	222 (100%)

χ² (df = 1, N = 222) = 7.8, p = .005.

H2d, which predicted that articles would be less likely to describe the Floyd incident as “police brutality” than the Nichols incident, was also supported. Articles covering Floyd’s killing only used “police brutality” an average of 1.56 times per article. Articles were significantly more likely to describe the Nichols killing as “police brutality” (M = 4.19). The difference, which was statistically significant, is shown in Table 7.

Table 7. Mean number of times act is described as “police brutality”.

Event	Mean Number of times act is described as “police brutality”
George Floyd	1.56 (SD = 1.8)
Tyre Nichols	4.19 (SD = 2.9)

Mean number of times act is described as “police brutality”: t(df = 221, N = 223) = 7.8, p < .001.

Sub-hypothesis H2e was the lone H2 sub-hypothesis to not be supported. The sub-hypothesis predicted that Fox News website articles would provide more personal details about police during coverage of the Floyd incident than during coverage of the Nichols incident. In fact, results ran opposite to expectations. More personal details were provided about police during coverage of the Nichols incident (M = 3.00) than the Floyd incident (M = 1.23). Table 8 shows the results.

Table 8. Mean number of personal details about police.

Event	Mean Number of Personal Details about Police
George Floyd	1.23 (SD = 2.2)
Tyre Nichols	3.00 (SD = 3.0)

Mean number of personal details about police: t(df = 163.7, N = 223) = 4.9, p < .001.

H2f predicted that Fox News articles would be less likely to suggest that police officers overstepped acceptable bounds against Floyd than Nichols. Results suggest support for this sub-hypothesis. While 65.6% of articles about Floyd did suggest that police had overstepped acceptable bounds in their treatment of him, this was significantly less than articles covering the killing of Nichols. A total of 87.4% of articles covering the Nichols incident suggested that police had overstepped acceptable bounds. As displayed in Table 9, a chi-square test showed the difference to be statistically significant [χ² (df = 1, N = 223) = 13.7, p < .001].

Table 9. Articles suggesting that police overstepped acceptable bounds.

	Articles suggesting that police overstepped acceptable bounds
Event	Articles suggesting that police overstepped	Articles not suggesting that police overstepped	Total
George Floyd	84 (65.6%)	44 (34.4%)	128 (100%)
Tyre Nichols	83 (87.4%)	12 (12.6%)	95 (100%)
Total	167 (74.9%)	56 (25.1%)	223 (100%)

χ² (df = 1, N = 223) = 13.7, p < .001.

The final H2 sub-hypothesis, H2g, predicted that articles published on the Fox News website would be more likely to suggest that police violence was justified against Floyd than against Nichols. Results, which are displayed in Table 10, show support for this sub-hypothesis. While only one 1.1% of articles about Nichols (N = 1) intimated that the violence carried out against him was justified, 10.9% of articles about Floyd (N = 14) justified the violence perpetrated against him. The difference was statistically significant at the .05 level [χ² (df = 1, N = 223) = 8.5, p = .004].

Table 10. Articles suggesting that police violence was justified.

	Articles suggesting that police violence was justified
Event	Articles suggesting that police violence was justified	Articles not suggesting that police violence was justified	Total
George Floyd	14 (10.9%)	114 (89.1%)	128 (100%)
Tyre Nichols	1 (1.1%)	94 (98.9%)	95 (100%)
Total	15 (6.7%)	208 (93.3%)	223 (100%)

χ² (df = 1, N = 223) = 8.5, p = .004.

This study also posed a pair of research questions. RQ1 asked about how Fox News talked about anti-police protests during its website’s coverage of the Floyd and Nichols incidents, respectively. The question asks, in particular, about whether a more critical tone would be adopted against anti-police protests following the Floyd incident.

Consistent with the direction of the general expectations of the research, findings suggest a more critical tone was adopted against anti-police protests following the Floyd incident. Overall, in the aftermath of the killing of Floyd, 51.6% of articles about anti-police protests carried an unfavorable tone about police. This compares with just 5.3% of articles following Nichols’ killing. It is also worth highlighting that more than 82% of articles about the Nichols incident did not make any mention of anti-police protests. This is unsurprising given the reality that Nichols’ killing did not trigger a large protest movement. A chi-square test showed statistically significant differences between the coverage of the two events. Table 11 shows the findings.

Table 11. Mention of anti-police protests.

	Mention of anti-police protests
Event	Favorable mention	Unfavorable mention	No mention	Total
George Floyd	31 (24.2%)	66 (51.6%)	31 (24.2%)	128 (100%)
Tyre Nichols	12 (12.6%)	5 (5.3%)	78 (82.1%)	95 (100%)
Total	43 (19.3%)	71 (31.8%)	109 (48.9%)	223 (100%)

χ² (df = 2, N = 223) = 77.9, p < .001.

RQ2 asked the following: What is the overall impression provided by Fox News about police in general during coverage of both incidents? Will the website’s coverage be more sympathetic toward police in the context of the Floyd incident?

Also consistent with general expectations, results here suggest more sympathy with police in the context of the Floyd incident than the Nichols incident. For the Floyd incident, a total of 7.8% of articles left an overall positive impression of police, compared with 78.1% of articles which left an overall negative impression and 14.1% of articles which were more neutral. These findings stand in stark contrast to articles covering the killing of Nichols. For the Nichols incident, only 1.1% of articles left a positive impression of police, compared with 98.9% of articles leaving an overall negative impression. No articles were neutral. Results, which show statistically significant differences, are displayed in Table 12.

Table 12. Overall impression of police.

	Overall impression of police
Event	Positive	Negative	Neutral	Total
George Floyd	10 (7.8%)	100 (78.1%)	18 (14.1%)	128 (100%)
Tyre Nichols	1 (1.1%)	94 (98.9%)	0 (0.0%)	95 (100%)
Total	11 (4.9%)	194 (87.0%)	18 (8.1%)	223 (100%)

χ² (df = 2, N = 223) = 21.1, p < .001.

Discussion

The goal of this quantitative content analysis was to examine the Fox News website framing of the police killings of two Black American males, George Floyd in May 2020 and Tyre Nichols in January 2023, respectively. A central goal of this study was to find out if there were discrepancies in the website’s framing of the two killings in light of the differences in the perpetrators’ races—A white police officer in Floyd’s murder and black officers in the case of Nichols.

As the literature review showed, many studies have been conducted analyzing U.S. media framing of race, including several studies examining Fox News coverage of race-related matters in general and Black people in particular. However, this study is the first to analyze two incidents of flagrant police brutality captured on video perpetrated against African-Americans by officers of different races. To the best of our knowledge, no previous research has comparatively examined the framing of brutality committed by black and white police officers in America.

The study was set up in a way that allowed the researchers to shed light on how the Fox News website delivered its frames of the two killings, and how it utilized devices to drive frames, including the use of keywords and expert testimonies.

Eight out of this study’s ten sub-hypotheses were supported. Moreover, findings from the two research questions fell in line with the predictions made in the sub-hypotheses. Despite the graphic brutality of both incidents (and their accompanying video footage), the website’s framing reflected striking differences between the two killings. Some of these differences were exemplified in the framing of the victims. The Fox News website was more sympathetic to Nichols than it was to Floyd. There was more humanization of Nichols than Floyd, something pointed to by the inclusion of more quotes and paraphrases from Nichols’ family members and friends as well as the inclusions of more personal details about him than Floyd. This raises the question of why the two dead Black males, both members of a nonwhite, less domineering “outgroup” (van Dijk, 2000), were not humanized to the same degree.

There were even more stark differences in the website’s framing of the perpetrators in the two killings. Overall, the site reflected a more sympathetic framing to the police officers involved in the Floyd killing than their peers involved in the Nichols killing. Findings pointed to significant differences in the humanization of the perpetrators in each killing. While there were quotes and paraphrases from police officers in both incidents, there were more quotes and paraphrases from police officers involved in the Floyd incident. This is an important finding, in part because, as the section on Framing Theory notes, sources can often drive frame direction (Entman, 1993; Gonçalves, 2023).

A more striking difference was in the number of expert condemnations of the police force, which tilted strongly toward the Nichols incident. Slightly more than 25% of the Nichols-related articles included expert condemnations compared to only 11% of Floyd-related articles.

A prominent framing difference was also reflected in the framing of police officers. A total of 14% of articles about Floyd framed police as honorable crime fighters, while none of the articles about the Nichols incident framed officers in this way. The notion of “police brutality” was also mentioned much more often in articles about the Nichols incident (M = 4.2) than the Floyd incident (M = 1.6).

Both videos featured excessive, graphic police violence against unarmed Black men. Articles covering the Floyd incident (65.6%), however, were much less likely than articles covering the Nichols incident (87.4%) to suggest that police had overstepped acceptable bounds of behavior. In the same context, close to 11% of the articles about Floyd justified the violence used against him, compared to only 1% of articles about Nichols.

Overall, the Fox News website was, to a certain extent, critical of police and relatively sympathetic with victims in its coverage of both the Floyd and Nichols killings. This may be surprising given past scholarly research into Fox News, but can arguably be explained by the explicitly graphic and particularly brutal nature of the content of both videos. In some ways, news outlets had no choice but to be critical of the police. At issue here, though, is the clear framing discrepancies that exist between articles about the two events, and the fact that Fox seemed to paint a comparatively rosy picture of the Floyd police killing. This is consistent with this study’s expectations, which are based on previous literature pointing to Fox News’ adoption of an ideology that endorses and reflects white supremacy (Mills, 2017). The core of Fox News’ functioning rests on “shoring up the power structures that continue to privilege white America” (Killen, 2018, p. 633).

The differences in the framing of the police officers in the two incidents can be further contextualized in the bipolarity of a strong, white ingroup and a weaker, nonwhite outgroup that was alluded to in the literature (Brewer, 1999; van Dijk, 2000). This bipolarity along racial lines is what might have made the Fox News website reflect a more sympathetic frame of police in close to 8% of articles about the Floyd incident compared to only 1% of articles about the Nichols incident. Prejudices and racial biases in favor of members of the more dominant, white ingroup can be attributed to “infra-humanization”, which has been defined as “a process by which people consider their ingroup as fully human and outgroups as less human and more animal-like” (Leyens et al., 2007, p. 140).

These findings are especially important in light of Fox News’ seemingly great impact on viewer perceptions on issues of race and racism in America (Jensen & Karell, 2023). This impact has critical implications for the overall fabric of American society, which continues to be meaningfully divided along racial lines. This unhealthy divisiveness becomes even more serious when incidents that are allegedly driven by racist motivations involve a perpetrator(s) from the establishment (the police in the case of this study) and a victim(s) from a vulnerable racial minority.

This study provides valuable insights about the content of Fox News website content. Future research should analyze the content of broadcast coverage and examine the potential impact of pro-police, pro-white, anti-black frames on Fox News viewers.

Disclosure statement

No potential conflict of interest was reported by the author(s).