MEDIA FRAMING, ISSUES, AND THE RACE OF PROTESTERS

Existing evidence suggests that media frames shape citizens’ orientation toward protest events. Citizens express less tolerance toward protests, particularly from socially vilified groups (e.g., the Ku Klux Klan) when presented with media frames that emphasize social order more than freedom of speech (Nelson, Clawson, and Oxley Reference Nelson, Clawson and Oxley1997). Moreover, the framing of social policies around concerns about crime and maintaining “law and order” have been shown to activate underlying racial bias among white Americans (Mendelberg Reference Mendelberg2001). Recent evidence demonstrates that media coverage framed around either concerns for “civil rights” or “social order” influences responses from the mass public and elites to Black protest (Wasow Reference Wasow2020; see also Kilgo and Mourão Reference Kilgo and Mourão2021; Reid and Craig Reference Reid and Craig2021). Furthermore, frames that emphasize the experiences of marginalized subgroups within the Black community have been shown to demobilize Black Americans who are not part of the subgroup (Bonilla and Tillery, Jr. Reference Bonilla and Tillery2020).

Beyond the media framing of protests, certain issues are more polarizing than others and therefore are more likely to draw public opposition. For example, there is considerable evidence that race has had a profound impact on American politics—particularly in the twentieth century—and can polarize public opinion on racial issues as well as seemingly race-neutral issues (Carmines and Stimson Reference Carmines and Stimson1989; Mendelberg Reference Mendelberg1997; Tesler Reference Tesler2012). There is evidence that sociocultural and moral issues similarly polarize the American public (Evans Reference Evans2003).

In addition, the race of protesters is likely to shape support for thwarting protest activity. Social protest serves as a way for aggrieved, typically socially marginalized groups to increase the salience of their grievances, have their issues placed on the institutional agenda, and challenge oppressive structures (Buechler Reference Buechler1993). There is a long tradition of civil rights protests among racial and ethnic minority groups in the United States, from the Black civil rights movement in the 1950s and 1960s, to the nationwide marches in support of immigration reform in 2006 (Voss and Bloemraad Reference Voss and Bloemraad2011), to the recent wave of Black Lives Matter protests in opposition to police shootings of unarmed Black citizens.

Although racial minority groups often rely on social protest to achieve institutional reform, polls show traditionally low levels of support for their exercise of free speech and assembly (Phoenix Reference Phoenix2020). Recent research reveals that when protesters are described as “all Black,” the protests are more likely to be perceived as potentially violent than when the protesters are described as “all white,” even when the former are depicted as nonviolent (Peay and Camarillo Reference Peay and Camarillo2021). Furthermore, evidence shows that Black protests garner more police action than predominantly white protest events (Davenport, Soule, and Armstrong Reference Davenport, Soule and Armstrong2011). The derogation of Black protest has been attributed, in part, to media frames that privilege whiteness and minimize Black grievance and modern protest tactics, typically by contrasting current Black protests with a sanitized narrative of the Black civil rights movement in the 1950s and 1960s (Jackson Reference Jackson2021). This tendency is exacerbated by the traditional media model, which gravitates toward narratives that dramatize and oversimplify complex social events (García-Perdomo, Harlow, and Brown Reference García-Perdomo, Harlow and Brown2022). Outside of the domain of immigration, little research has explored the public response to protests with predominantly Latina/o protesters (Branton et al. Reference Branton, Martinez-Ebers, Carey and Matsubayashi2015; Carey, Branton, and Martinez-Ebers Reference Carey, Branton and Martinez-Ebers2014; Wallace, Zepeda-Millán, and Jones-Correa Reference Wallace, Zepeda-Millán and Jones-Correa2014; Zepeda-Millán Reference Zepeda-Millán2017). This study advances the existing literature by comparing reactions to protest events among white, Black, and Latina/o protesters.

Finally, citizens may be less inclined to suppress protests from groups that share their own racial and ethnic background. Considerable evidence demonstrates that groups tend to mobilize around issues that appear to advance group interests (Dawson Reference Dawson1995; Sanchez and Masuoka Reference Sanchez and Masuoka2010). Moreover, group members are “group-centric” in their policy support, which means that their support for policies depends on the group that is perceived to benefit the most from the policy (Nelson and Kinder Reference Nelson and Kinder1996). Although the literature suggests that this increases opposition toward policies that are perceived to benefit unfavorable groups, group-centrism also implies that group members will support policies that are perceived to benefit in-group members.

Given the existing literature, our expectations were as follows:

DATA AND METHODS

To test our hypotheses, we used data from a population survey funded by the Knight Foundation to examine attitudes toward free speech and expression. The survey was fielded in the summer of 2021, and the sample included 5,299 US adults recruited from an online, probability-based Ipsos KnowledgePanel.® Participants were given the option to complete the survey in English or Spanish. One benefit of the sample was that it included oversamples of 941 non-Hispanic Black respondents and 967 Latina/o respondents in addition to 2,443 non-Hispanic white respondents. The remaining respondents identified with other non-Hispanic racial groups or as multiracial. The substantial oversamples of Black and Latina/o respondents allowed for a more accurate assessment of how their attitudes are shaped by the experimental treatments. The analysis for this article was conducted on a split sample within the full sample of respondents. The split sample consisted of 2,411 respondents and included 1,125 non-Hispanic white respondents, 440 non-Hispanic Black respondents, and 393 Latina/o respondents (Carey and Cisneros Reference Carey and Cisneros2022).

Another benefit of the survey was that we were allowed to embed an experiment that explores the influence of the framing of the protest, the issues, and the race of protesters. As a test of our key hypotheses, we conducted a preregistered experiment (see https://osf.io/6bcuf, EGAP Registration ID: 20210707AAA) with a between-subjects, 2 (i.e., media frame: free speech or social order) x 3 (i.e., race of protesters: white, Latina/o, or Black) x 2 (i.e., issue: policing or the environment) factorial design. Participants were assigned randomly to a condition in which they read a faux newspaper article that described a protest event. Half of the conditions used a frame that stressed the protest as a potential threat to social order; the other half used a frame that emphasized protesters’ right to exercise their free speech. The manipulation was done in the headline. The social-order media frame read: “Mayor Vows to Keep Law and Order”; the free-speech frame was: “Mayor Supports Protesters’ Right to Protest.” The text of the article applied these distinct frames in several parts. First, in the description of the local police response, the social-order frame stated: “The police department brought in more officers to secure buildings and the city and county imposed a curfew.” In contrast, the passage in the free-speech frame stated: “The police department released a press release supporting the right of the protesters to voice their concerns and shared there would be a light police presence during the event.” Later, the social-order frame attributed a statement to the mayor in which he stated: “I intend to do everything in my power to ensure public safety. We will arrest and prosecute anybody who is engaged in disorderly conduct.” Alternatively, in the free-speech frame, the mayor stated: “Peaceful protesters deserve the space and protections to make their voices heard.”

The second factor in the experiment was the race of the protesters, described as either predominantly white, Latina/o, or Black. In addition, images of protesters were included in each treatment that matched the racial and ethnic backgrounds of the descriptions. To facilitate fair comparisons across each image, we chose images in which (1) protesters were actively protesting, (2) protests were visibly nonviolent, and (3) there was limited signage.Footnote ¹ If signs were included in the image, they were in the background and protest-related so that potential confounds were not introduced into the treatment. We also included a third dimension for our experiment that manipulated whether the protest described in the treatment concerned policing or the environment. Some participants also were assigned to a control condition that did not expose them to any content. They were directed to answer the key dependent variable.

After exposure to the experimental treatments, participants were asked to respond to the key dependent variable, which measured support for police suppression of protests. Respondents were asked how much they agree with the following statement: “Sometimes police need to use force to suppress protesters.”Footnote ² They responded to the question using a 10-point scale ranging from 1 (strongly disagree) to 10 (strongly agree). We used ordinary least squares regression to test our hypotheses.