INTRODUCTION
Black Lives Matter
This declaration, originally conceived in 2013 after George Zimmerman's acquittal for the death of Trayvon Martin and used as a rallying cry after the shooting of Michael Brown in Ferguson, Missouri, has come to represent a larger social movement reflected both in mass protests over a series of deaths of Black citizens at the hands of police and in the work of activists and organizers across the United States (e.g., Reference LoweryLowery, 2016).Footnote 1 The declaration signifies the root concern of the movement: the devaluation of Black lives evidenced in broad racial inequalities and injustices and highlighted specifically in the disproportionate experience of hostile or fatal interactions between police and Black civilians.Footnote 2
Research suggests the movement's grievances are well-founded. Broadly, Black Americans experience persistent and striking socioeconomic inequalities and disparate exposure to the criminal justice system (e.g., Reference Peterson and KrivoPeterson & Krivo, 2010). More specifically, the police do, on average, treat Black citizens differently (e.g., Reference Baumgartner, Epp and ShoubBaumgartner et al., 2018; Reference Crutchfield, Skinner, Haggerty, McGlynn and CatalanoCrutchfield et al., 2012; Reference Gelman, Fagan and KissGelman et al., 2007; US DOJ Report, 2013; Reference Voigt, Camp, Prabhakaran, Hamilton, Hetey, Griffiths, Jurgens, Jurafsky and EberhardtVoigt et al., 2017), including shooting and killing unarmed Black people more frequently than unarmed white people (e.g., Reference Correll, Park, Judd, Wittenbrink, Sadler and KeeseeCorrell et al., 2007; Reference Nix, Campbell, Byers and AlpertNix et al., 2017; Reference RossRoss, 2015).
Despite this, most white Americans do not express support for the Black Lives Matter (BLM) movement. In 2016, only 40% of white Americans expressed any kind of support for BLM and only 14% expressed strong support (Pew Research Center, 2016).Footnote 3 Only 28% of white Americans thought that Trayvon Martin's death raised important issues about race, and only 37% felt the same about Michael Brown's shooting (Pew Research Center, 2014). This is in stark contrast to Black Americans, who largely thought both cases raised important questions about race (around 80% for each) and 65% of whom supported BLM, with 41% strongly supportive (Pew Research Center, 2014, 2016).Footnote 4 Even among police officers, Black officers are substantially more likely than white officers to believe that the deaths of Black citizens at the hands of the police are signs of a broader problem (Pew Research Center, 2017).
Why is it that white Americans hold substantially less favorable views of the movement than Black Americans? To answer this question, we focus on a particular moment in time: the crest of the first wave of widespread attention to the movement in 2016, when protests were widespread across the country and the movement became a significant issue in the presidential election. Based on the specific and general concerns of the movement—police killings of Black Americans as a window into broader racial inequalities— and drawing on theory and past research, we develop and test four potential explanations for the divergence in perspectives. Two are at least superficially nonracial: that differences in support for the movement might be explained by racial disparities in experiences with the police or more generally in differences in affinity for the police. The other two propose a role for racism in explaining views of this racial civil rights movement, investigating basic racial affinity as well as a more modern form of racism that implicates the movement's broader concern: the relative positions of Black and white people in America's racial hierarchy, and the threat to the current racial order posed by the movement.
It matters that so few white Americans feel positively toward BLM. More than 150 years after the end of chattel slavery and more than 50 years after a racial civil rights movement challenged legal segregation and discrimination, the United States remains deeply divided about the persistence of racial inequality and injustice. Public opinion about the central concerns of the movement had political relevance in the 2016 election (e.g. Reference Drakulich, Hagan, Johnson and WozniakDrakulich et al., 2017, Reference Drakulich, Wozniak, Hagan and Johnson2020), and of course has broader consequences for the future of the country's racial divide.
For these same reasons, it also matters why white Americans are less supportive of the movement. Scholarly work generally frames opposition to the civil rights movement of the 1950s and 1960s as a function of racism (e.g., Reference Beckett and SassonBeckett & Sasson, 2004; Reference Sears and KinderSears & Kinder, 1971; Reference Sears and McConahaySears & McConahay, 1973; Reference TonryTonry, 2011), suggesting opposition to the BLM movement may be similarly explained. However, there are two related reasons to consider alternative explanations. First, competing public narratives about the motivations behind support or opposition to BLM compel a systematic evaluation. Specifically, counter-protestors and social media critics argued that their opposition was rooted in their support for the police, and explicitly denied racism as a motivation.Footnote 5 Second, real differences exist between the 1960s and 2010s, including a striking decline in overt expressions of racism (Reference Krysan and MobergKrysan & Moberg, 2016; Reference Schuman, Steeh, Bobo and KrysanSchuman et al., 1997). As a result, it is important to carefully examine the nature and sources of attitudes toward BLM.
The explanations we consider each imply different causal mechanisms. A difference in experiences with the police highlights a white ignorance of the experiences of Black Americans, perhaps protected by persistent segregation. Affinity toward the police speaks to competing claims—that feelings about BLM are not fundamentally about race, but about support for the symbolic guardians of the moral order, or, on the other side, about a hatred of those same guardians and the order they represent (an accusation made by the Blue Lives Matter counter-movement). The explanation rooted in an affinity for Black people reflects the problem of affective ties too rarely reaching across racial lines. Or opposition to the movement might be better explained by a more modern racism—a perspective that anticipates claims of superficially nonracial motivations. This implies something much more problematic: that white opposition to the movement reflects a more fundamental opposition to efforts to upset the racial status quo.
We conclude by situating our findings about views of BLM in historical context, drawing both parallels and distinctions between the Civil Rights Era and the modern one. In the end, racial and racist progress appears to run in parallel and interactive rather than pendular and divergent paths (e.g., Reference KendiKendi, 2016). We also consider the implications of our findings for understanding attitudes toward the movement beyond 2016.
FOUR EXPLANATIONS FOR RACIAL DIFFERENCES IN VIEWS OF BLM
We begin by developing four theoretical explanations for why white Americans, on average, feel substantially less favorably toward BLM than Black Americans.Footnote 6 While the four explanations are not necessarily competing—each could explain some of the white-Black difference—we do not present them as equally likely, noting in particular serious questions about the first three that may be resolved by the fourth. The first two explanations are ostensibly race-neutral and focus on experiences with and feelings toward the police. The latter two focus on the role of racism, distinguishing more overt expressions of racial animus from more strategic concerns about group position (e.g., Reference BlumerBlumer, 1958).
The police
The first possible explanation is that divergent views of a movement concerned with police misconduct are a consequence of race differences in police contact (Reference Braga, Brunson and DrakulichBraga et al., 2019). People of color—and people in communities of color—experience a greater quantity of police contact, including stops and frisks, searches, citations, and arrests (e.g., Reference Baumgartner, Epp and ShoubBaumgartner et al., 2018; Reference Crutchfield, Skinner, Haggerty, McGlynn and CatalanoCrutchfield et al., 2012; Reference Fagan and DaviesFagan & Davies, 2000; Reference Gelman, Fagan and KissGelman et al., 2007; Reference IngramIngram, 2007; Reference Kochel, Wilson and MastrofskiKochel et al., 2011; Reference LytleLytle, 2014; Reference Petrocelli, Piquero and SmithPetrocelli et al., 2003; Reference Rojek, Rosenfeld and DeckerRojek et al., 2012). While white Americans are more likely to initiate contact with the police or have contact with officers offering help, Black Americans are more likely to be stopped while driving and much more likely to be stopped and questioned on the street (e.g., Reference Davis and WhydeDavis & Whyde, 2018). Within this contact, people of color—and people in communities of color—are also more likely to experience police disrespect and misconduct (e.g., Reference Epp, Maynard-Moody and Haider-MarkelEpp et al., 2014; Reference KaneKane, 2002; US DOJ, 2013; Reference Voigt, Camp, Prabhakaran, Hamilton, Hetey, Griffiths, Jurgens, Jurafsky and EberhardtVoigt et al., 2017), the use of coercive force (e.g. Reference Smith, Reiss and TonrySmith, 1986; Reference Terrill and ReisigTerrill & Reisig, 2003), and be injured, shot, or killed (Reference Correll, Park, Judd, Wittenbrink, Sadler and KeeseeCorrell et al., 2007; Reference Nix, Campbell, Byers and AlpertNix et al., 2017; Reference RossRoss, 2015; Reference Ross, Winterhalder and McElreathRoss et al., 2018).
In judging the likelihood of a phenomenon, people tend to privilege their personal experiences (e.g., Reference Tversky and KahnemanTversky & Kahneman, 1974). Persistently high levels of racial segregation in the residential, professional, and social lives of Americans (Reference Massey and DentonMassey & Denton, 1993; Reference Peterson and KrivoPeterson & Krivo, 2010) insulate many white Americans from the treatment many Black Americans receive at the hands of the police. Consequently, views of the police appear to be stratified not just by the race of individuals but also by the racial composition of the neighborhoods in which Americans reside (e.g., Reference DrakulichDrakulich, 2013; Reference Drakulich and CrutchfieldDrakulich & Crutchfield, 2013). And although raising awareness about these kinds of experiences was an explicit goal of the BLM movement, people's selection into media outlets—themselves relatively segregated—likely prevented many Americans from learning about these experiences in a manner that does not simply reinforce pre-existing beliefs. In short, white Americans may not support BLM both because of their own lack of experience with aggressive police stops as well as their social distance from those who have experienced frequent police stops.
One important problem clouds this explanation. Although people may privilege their own experiences in the absence of direct knowledge of the experience of Black Americans, they also bear some responsibility for this lack of knowledge given the choices they made that resulted in a lack of direct contact with Black Americans or in the media outlets they seek out. Even when people do encounter stories of the problematic treatment of Black Americans by police, they may choose not to believe them or interpret them as exceptional rather than systematic.Footnote 7 In this light, an explanation resting on benign ignorance seems insufficient. Reference MillsCharles Mills (1997, Reference Mills, Sullivan and Tuana2007) describes a white ignorance—protected by an “epistemology of ignorance”—that acts to intentionally insulate white Americans from the realities of race. This willful ignorance of inequalities is a core component of theories of modern racism, including laissez-faire (e.g., Reference Bobo, Kluegel, Tuch and MartinBobo & Kluegel, 1997; Reference Bobo, Smith, Katkin, Landsman and TyreeBobo & Smith, 1998) and color-blind racism (Reference Bonilla-SilvaBonilla-Silva, 2017), as we discuss below. In the context of understanding disparities in the criminal justice system, Reference Drakulich, Rodriguez-Whitney, Hollis, Martinez and StowellDrakulich and Rodriguez-Whitney (2018: 23) describe this as a failure “to critically investigate the inconvenient truths of racial privilege.”
The second possible explanation for Black-white differences in views of the movement is that opposition to the movement is driven not just by experiences with the police but more broadly by feelings toward the police. The police act as representatives of the legal authority and are publicly tasked with protecting citizens from criminal harm and providing order. Psychologists identify “respect for authority” as one of the core moral foundations: nearly everyone has some level of respect for authority although people differ in how important this moral foundation is relative to others like fairness or caring (Reference Graham, Haidt, Koleva, Motyl, Iyer, Wojcik and DittoGraham et al., 2012; Reference Haidt and JosephHaidt & Joseph, 2004). Thus, many people may have an affinity for the police, especially when they see them as providing important social goods like security and order. This may be especially true amidst broader concerns about social changes in which people may “look to the police to defend a sense of order” (Reference Jackson and BradfordJackson & Bradford, 2009, p. 499; see also Reference Jackson and SunshineJackson & Sunshine, 2007; Reference WozniakWozniak, 2016). If there are racial differences in support or affinity for the police, they may explain differences in views of BLM. In fact, BLM protestors and politicians who supported them were accused of lacking the proper respect for the police, and even of being motivated by a dislike or hatred of the police (e.g., Reference AlcindorAlcindor, 2016; Reference HolleyHolley, 2016; Reference VitaleVitale, 2016).
To the extent that there are racial differences in support for the police, it is also important to consider why they exist, as this may complicate this potential explanation. Work on moral foundations often specifies that this respect is for a legitimate authority (e.g. Reference Graham, Haidt and NosekGraham et al., 2009). The relationship between legitimacy and respect for authority may be tautological: legitimacy is predicated on the existence of respect for authority and this respect only makes sense when the authority is viewed as legitimate. In this sense, it matters that many Americans, particularly people of color, have sincere questions about the legitimacy of the police (e.g., Reference Peffley and HurwitzPeffley & Hurwitz, 2010). Several theories take different approaches to this phenomena, describing the concerns as rooted in procedurally unjust experiences with the police (e.g., Reference TylerTyler, 1990; Reference Tyler and WaslakTyler & Waslak, 2004), as a legal cynicism in which the “the police and courts are viewed as illegitimate, unresponsive, and ill equipped to ensure public safety” (Reference Kirk and PapachristosKirk & Papachristos, 2011, p. 1191), or as a legal estrangement in which many in poor communities of color correctly recognize the law as operating to exclude them from society (Reference BellBell, 2017). If the police frequently act in racially biased ways, this not only recasts lower levels of support among Black Americans as justified, it also raises troubling questions about the meaning of white support. In short, what does it mean to support the police when the police are publicly accused of racial bias?
Racism
The third possible explanation for interracial differences in attitudes toward BLM is that feelings about the movement are rooted in feelings about Black people. This is among the simplest explanations and reflects the logic of the name of the movement, implying that Black lives are not valued equally by all Americans. Research on the contact hypothesis (Reference AllportAllport, 1954) suggests positive intergroup feelings emerge from positive contact between members of different groups (Reference PettigrewPettigrew, 1998; Reference Pettigrew and TroppPettigrew & Tropp, 2006). Unfortunately, high levels of segregation, especially in residential communities (e.g. Reference Peterson and KrivoPeterson & Krivo, 2010), mean that affective ties often fail to stretch across racial lines. In this simple sense, then, a greater affinity for Black people among Black people may help explain their greater affinity for BLM.
There is a long history of anti-Black animus among white Americans, one that justified slavery (e.g., Reference KendiKendi, 2016), the post-Civil War Black Codes (e.g. Reference BlackmonBlackmon, 2008), the Jim Crow System (e.g. Reference AlexanderAlexander, 2010), and even biased policies in northern cities during the Progressive Era (Reference MuhammadMuhammad, 2019). Thus, antipathy toward BLM may simply be a function of antipathy toward Black people. However, this explanation faces two related potential problems. First, relatively explicit expressions of racial animus—like a lack of warm feelings toward Black people as a group—have declined substantially (Reference Krysan and MobergKrysan & Moberg, 2016; Reference Schuman, Steeh, Bobo and KrysanSchuman et al., 1997). Second, many theories of racism focus on group interest rather than simple affection (e.g. Reference BlumerBlumer, 1958; Reference Bonilla-SilvaBonilla-Silva, 2017)—something that was true even when overt expressions of racism were more common (e.g., Reference KendiKendi, 2016). Further, despite the seeming decline in overt expressions of racial animus, support for policies that disproportionately harm Black Americans and opposition to policies designed to ameliorate racial inequalities suggest a more modern form of racism not reflected in simple measures of racial animus (e.g., Reference Bonilla-SilvaBonilla-Silva, 2017; Reference DrakulichDrakulich, 2015a, Reference Drakulich2015b). Thus, a simple indicator of racial feelings may miss a more complicated racial logic that better explains opposition to BLM.
The civil rights movement and subsequent decline in overt racism motivate the fourth and final possible explanation, which focuses not on a simple view of racial affect but on perceptions of racial group interest. A civil rights movement, by definition, presents a threat to the status quo racial hierarchy. Any perceived threats to the existing racial order and existing distribution of privileges and resources will engender racial prejudice as a protective mechanism (e.g. Reference BlumerBlumer, 1958). This can be seen most dramatically in the rise of racist collective action—including by the Ku Klux Klan (see Reference CunninghamCunningham, 2013; Reference McVeighMcVeigh, 2009)—to preserve racial privileges thought to be threatened, but may also be reflected more broadly in public opinion among white Americans. Similar to the lack of support for the BLM movement, a majority of Americans—and white Americans in particular—opposed the protests of the 1960s civil rights movement (Reference LegumLegum, 2017).
In fact, “modern” theories of racism emerged out of attempts to understand the evolving views of race relations among white Americans both during the Civil Rights Era and in the decades following. Using a survey of Los Angeles residents collected in the months following the Watts Rebellion in 1965, symbolic racism was developed to help understand an emerging phenomenon in which some white Americans expressed support for racial equality in principle but simultaneously opposed policies attempting to ameliorate this inequality (Reference Kinder and SearsKinder & Sears, 1981; Reference Sears and KinderSears & Kinder, 1971; Reference Sears and McConahaySears & McConahay, 1973). During the Jim Crow Era, some white Americans used openly racist ideologies to justify their dominant position in the racial hierarchy, arguing that Black Americans were fundamentally inferior and thus undeserving of the same legal status as white Americans (e.g., Reference KendiKendi, 2016). However, when legal discrimination and racial inequalities were challenged—as they were by the civil rights movement—those in power began to shift away from openly racial ideologies and toward those emphasizing individualism while simultaneously minimizing or denying the role of historical or contemporary racism as barriers faced by Black Americans (Reference Jackman and MuhaJackman & Muha, 1984; Reference KrysanKrysan, 2000; Reference MendelbergMendelberg, 2001). This new form of racism, more covert but still effective at maintaining racial inequalities, has also been described as laissez-faire racism (Reference Bobo, Kluegel, Smith, Tuch and MartinBobo et al., 1997; Reference Bobo, Smith, Katkin, Landsman and TyreeBobo & Smith, 1998) or color-blind racism (Reference Bonilla-SilvaBonilla-Silva, 2017). A substantial body of prior research has connected these kinds of ideologies to opposition to economic policies intended to address racial inequalities (see Reference Bonilla-SilvaBonilla-Silva, 2017; Reference KrysanKrysan, 2000; Reference Tuch and HughesTuch & Hughes, 2011), while a smaller number of studies have connected them to support for criminal justice policies and practices that disproportionately harm Black Americans (e.g., Reference Carter and CorraCarter & Corra, 2016; Reference JohnsonJohnson, 2008), Thus, while some of the opposition to a racial civil rights movement like BLM may be expressed in openly racial and racialized terms, some may be expressed in color-blind or laissez-faire terms: emphasizing that existing inequalities are the product of individual successes and failures while intentionally ignoring the very different circumstances faced by members of different racial groups.
This fourth potential explanation addresses some of the limitations of the three preceding explanations. First, we suggested that white ignorance about the experience of Black Americans with the police was an unsatisfying explanation when that ignorance is willful and instrumental as color-blind and laissez-faire theories of modern racism suggest (Reference Bobo, Kluegel, Tuch and MartinBobo & Kluegel, 1997; Reference Bonilla-SilvaBonilla-Silva, 2017). Second, we asked what support for the police meant in the context of accusations of racial bias against the police. Theories of modern racism suggest people may mask racist motivations—whether possessed explicitly or implicitly (Reference DrakulichDrakulich, 2015a, Reference Drakulich2015b)—in superficially nonracial positions (Reference Bobo, Kluegel, Smith, Tuch and MartinBobo et al., 1997; Reference Bonilla-SilvaBonilla-Silva, 2017). In other words, some white people may justify their opposition to BLM on the grounds of support for the police when their opposition is really rooted in indifference toward Black lives and an opposition to calls for racial justice. Finally, the fourth explanation also accounts for the changing nature of racial prejudice from more overt anti-Black bigotry to more covert and implicit modern expressions of racism.
THE PRESENT STUDY
As a relatively new social movement—though one built on prior racial social justice movements (e.g., Reference LebronLebron, 2017; Reference LoweryLowery, 2016)—few studies have measured public opinion about BLM. One study examined the correlates of views of BLM and found that opposition was highest among older, conservative, and Republican men, as well as those who live in more Republican states (Reference Updegrove, Cooper, Orrick and PiqueroUpdegrove et al., 2018).
Despite the paucity of direct work on broad public opinion about BLM, research on the movement itself suggests processes consistent with our theoretical emphasis on group position. Reference BoylesBoyles (2019) describes a “post-Ferguson backlash against black action” (p. 29) that dismisses civil rights protest as disobedient and disorderly. Reference CobbinaCobbina (2019) situates the movement in the history of differential policing by race, linking these disparities to historical and contemporary processes of oppression. Notably, both of these scholars point to important variation in views of the movement among Black Americans.
The four explanations we explore are not necessarily conflicting or mutually exclusive, though we are interested in the relative magnitude of their effects. Based on the discussion above, we are particularly interested in the effects of the other explanations once we control for racial resentment, the measure capturing our fourth explanation.
Both major party presidential candidates in the 2016 election took stances on BLM. Hillary Clinton, the Democratic candidate, more closely aligned herself with the movement, calling for criminal justice reforms while also praising the police generally (Reference BlakeBlake, 2016; Reference BouieBouie, 2015; Reference GlantonGlanton, 2016; Reference GrawertGrawert, 2016). Donald Trump, the Republican candidate, explicitly criticized the movement, describing it as “trouble” (Campbell 2015) and suggested a protestor allegedly assaulted at one of his rallies “maybe … should have been roughed up” (Reference Johnson and JordanJohnson & Jordan, 2015). For this reason, it is important to control for the role of political identification.
However, the role of politics is complicated. Political identification is likely endogenous to racial resentment and racial animus (e.g., Reference DrakulichDrakulich, 2015a; Reference Matsueda, Drakulich, Hagan, Krivo, Peterson, Aldrich and McGrawMatsueda et al., 2011). Indeed, an explicit component of the Republican “southern strategy” was the recruitment of voters who were uncomfortable with the threats to the racial status quo posed by the Civil Rights Movement (e.g., Reference Beckett and SassonBeckett & Sasson, 2004; Reference TonryTonry, 2011). In the modern context, the nomination of a Black presidential candidate in 2008 appeared to push white voters with racial animus and resentments away from the Democrats (e.g., Reference Krupnikov and PistonKrupnikov & Piston, 2015). Support for the Tea Party Movement that arose shortly after this election was driven directly by racial resentment (e.g., Reference HochschildHochschild, 2016; Reference Tope, Pickett and ChiricosTope et al., 2015). In 2016, Donald Trump appeared to make direct appeals to these voters who felt left behind by structural economic changes and believed that racial minorities were “cutting the line” with the assistance of the federal government to gain the benefits of the American Dream (e.g., Reference Abramowitz and McCoyAbramowitz & McCoy, 2019; Reference Drakulich, Hagan, Johnson and WozniakDrakulich et al., 2017; Reference HochschildHochschild, 2016; Reference Sides, Tesler and VavreckSides et al., 2018). In fact, expressions of support for the police appeared to function as a racist dog whistle, predicting support for Donald Trump only among those high in racial resentment (Reference Drakulich, Wozniak, Hagan and JohnsonDrakulich et al., 2020). In short, political frames of the movement may have shaped public opinion both along party lines and in ways that resonated with those who have concerns about the relative status of Black and white Americans. This suggests that controlling for people's politics will allow us to examine the role of racism or views of the police independent of these political attitudes, but also has implications for how to interpret the meaning of the role of politics relative to our other explanations.
DATA, MEASURES, AND METHODS
We use data from the 2016 American National Election Studies Time Series Survey to explore racial differences in feelings toward the BLM movement. The survey was conducted in two modes: face-to-face interviews as well as questionnaires administered online (Reference DeBell, Amsbary, Meldener, Brock and MaiselDeBell et al., 2018).Footnote 8 The survey was conducted in two waves in late 2016—the first in the 2 months prior to the election and the second in the 2 months after the election. This timing works well to capture feelings about BLM at the crest of the first peak in widespread attention to the movement. Although the deaths of Black Americans provoked substantial and sustained protests in 2014 and 2015—particularly in Ferguson and Baltimore—the summer of 2016 saw these protests expand broadly in cities across the country (Reference Lee, Mykhyalyshyn, Omri and SinghviLee et al., 2016). Discussions of the movement by the major party candidates in the 2016 presidential election brought even more attention. Online searches suggest a first major peak of attention to the movement in the summer of 2016.Footnote 9 Our study captures the moment just after this first peak in news coverage and public interest in the movement. The post-election interview included 1059 face-to-face interviews and 2590 online interviews. The study is designed to be nationally representative of adult U.S. citizens.
Measures
Our core interest is in people's feelings toward BLM, which we measure with a “feeling thermometer scale,” a common format in political science research to gauge affective support for individuals, groups, or issues (Reference Nelson and LavrakasNelson, 2008). Respondents were asked to rate their feelings toward a list of individual persons and groups on a scale ranging from 0 to 100, with 0 representing very cold and unfavorable feelings and 100 representing very warm or favorable feelings. For this measure of warmth toward BLM, respondents were asked specifically how warmly or coldly they felt toward BLM. Overall, the average respondent was mixed in their feelings, rating the movement about a 50 on the scale from 0 to 100 (Table 1).Footnote 10
Table 1. Descriptive statistics
Note: N = 3252.
Abbreviations: BLM, Black Lives Matter; SD, standard deviation.
Two questions capture experiences with the police. The first asks whether the respondent or any of their family members were stopped or questioned by a police officer in the last 12 months. Roughly a quarter of respondents said yes. The second question asks whether the respondent has ever been arrested. Roughly a fifth of respondents reported they had.
In addition to experiences with the police, respondents were asked for their feelings about the police (warmth toward police) also using a thermometer scale. The average respondent felt relatively warmly toward the police, ranking them a 75 out of 100.
Similarly, respondents were asked about their warmth or coldness toward “Blacks” on a 100-point scale—representing explicit and affective views toward “Blacks” as a group. The average respondent felt somewhat warmly toward “Blacks”: a 68 out of 100.
Finally, concerns about the racial order were captured with a measure of racial resentment, a dimension of “symbolic racism” widely used in prior work (Reference Henry and SearsHenry & Sears, 2002; Reference Kinder and SearsKinder & Sears, 1981). These views are primarily driven by social concerns about relative racial group positions rather than by nonracial ideological views (Reference Simmons and BoboSimmons & Bobo, 2019). The measure is captured as the average of nonmissing responses to four questions (on five-point agreement scales): whether “Blacks” should overcome prejudice and work their way up without “special favors,” whether slavery and discrimination created conditions that remain significant barriers for lower-class “Blacks” (reverse-coded), whether “Blacks” had gotten less than they deserved (reverse-coded), and whether inequalities would be solved if “Blacks” tried harder.
We include two measures of political beliefs: identification as more liberal or conservative and as more Democrat or Republican, each on seven-point scales.
We control for a variety of sociodemographic characteristics that may be associated with views of this social movement: gender, age, education, income, employment status, race-ethnicity, and foreign-born. For race-ethnicity, we focus on the differences between non-Hispanic white and Black respondents, though also control for those identifying as Hispanic/Latino, Asian, or something else (including Native American and two or more races).Footnote 11
Methods
The study employs a stratified, clustered address-based sampling design for the face-to-face surveys and a somewhat simpler address-based sampling design for the online survey. The survey also includes weights to adjust for the sample design and attrition. To properly account for the survey weights within the stratified, clustered sampling scheme, the reported effects are estimated with survey-weighted generalized linear models using the “survey” package in R (Reference LumleyLumley, 2014; R Core Team, 2016). We test for mediation via a structural equation model using the “lavaan” and “lavaan.survey” packages (Reference OberskiOberski, 2014; Reference RosseelRosseel, 2012).
Relatively little data are missing within the survey—the only measures missing more than 2% of cases were income (4.2%) and age (2.6%). Recognizing that there is no perfect solution to missing data, we conducted all of the analyses twice, using different methods with different sets of assumptions. First and most simply, we drop the relatively small number of cases with missing data. Second, we employ a multiple imputation strategy (e.g., Reference AllisonAllison, 2002) using the “mice” package (Reference van Buuren and Groothuis-Oudshoornvan Buuren & Groothuis-Oudshoorn, 2011), imputing 20 data sets in a process that used all the variables from the analyses as well as several auxiliary variables to add information and increase efficiency. The auxiliary variables included information on marital status, children, home ownership, stock market investment, and length of time in the neighborhood. Reassuringly, the two approaches produced substantively identical results: no differences in statistical significance and similar magnitudes for the estimated effects. Given this, we report the simpler complete models.
Although the study is conducted in two waves, many of the key questions including feelings toward BLM and the police were asked only in the second wave, making the analyses effectively cross-sectional. Given the recent emergence of BLM, it seems likely that views of the police and race were established earlier, though we are unable to definitively establish the causal order.
RESULTS
Racial differences in experiences and views
Table 2 presents basic racial differences in support for BLM as well as for each of the potential explanations for this support. As expected, there are large racial differences in ratings of BLM. The average white respondent felt coldly toward the movement, rating it a 43 on a scale of 0–100.Footnote 12 The average Black respondent, by contrast, felt very warmly, rating it an 82, nearly 40 points higher than the average for white respondents.
Table 2. Racial differences in experiences and perceptions
Note: For tests of significance of difference from whites: ***p < 0.001, **p < 0.01, *p < 0.05.
Abbreviation: BLM, Black Lives Matter.
Consistent with prior work on police stops, sizeable interracial differences existed in reports that the respondent or a family member had been stopped or questioned by the police in the prior year. More than a one-third of Black respondents reported that they or a family member had been recently stopped or questioned by the police, as compared to less than a quarter of white respondents. Despite the large difference in recent stops, the proportions of Black and white respondents who had ever been arrested were roughly equal.
Feelings toward the police also differed greatly between Black and white respondents. White respondents rated police highly, an average of nearly 80 out of 100. Black respondents were much more mixed, rating the police on average as a 55 on the 0 to 100 scale. Importantly, while support for BLM was relatively consistently high among Black respondents—half of Black respondents rated the movement above 85 on the 100-point scale—there was substantial variation in feelings toward the police among Black respondents. Slightly more than half the sample felt positively toward the police, ranking them a 51 or higher with 30% ranking them 70 or above and 15% ranking them 85 or above. Seven percent ranked the police a perfect 100 while 5% ranked them 0.
Not surprisingly, there were also substantial differences in racial attitudes and views. On average, white respondents reported feeling only lukewarm toward “Blacks”: 67 out of 100. Black respondents, by contrast, felt very warmly toward other “Blacks”: an average of 82 out of 100. Most white respondents—57%—were at least a little racially resentful of Black Americans, and 27% were very racially resentful.Footnote 13 Levels of racial resentment were much lower for Black respondents.
Explaining differences in feelings toward BLM
Table 3 presents four basic models to explore people's feelings about BLM. The first includes basic controls. The second explores the two explanations related to the police: experiences and affinity. The third explores the role of racial feelings and views: an affinity or animus toward Black people and racial resentment. The final includes controls for political identification.Footnote 14 We pause in discussing the second and fourth models to consider potential moderating effects for key variables. We then conclude with a more formal test of whether any of our explanations mediate the racial differences in views between Black and white Americans.
Table 3. Coefficients from regression of support for Black Lives Matter
Note: N = 3252.
Abbreviations: β, unstandardized coefficient; AIC, Akaike information criterion; SE, standard error; Zβ, standardized coefficient.
***p < 0.001, **p < 0.01, *p < 0.05.
Model 1 includes basic controls for demographic and biographical characteristics. On average, female respondents rated BLM about nine points more warmly than male respondents. Those currently or formerly married rated BLM less warmly than single respondents, while those with more years of education rated the movement more favorably. Even after accounting for differences in these basic characteristics, Black and white respondents were still 38 points apart in their average ratings of the movement (the coefficient for Black represents the average difference in ratings between Black and white respondents). Asian respondents were not significantly different from whites in their assessment of BLM, while those identifying themselves as Hispanic or Latino were significantly more supportive of BLM than white respondents.
Role of experiences with and feelings toward the police
Model 2 adds measures to explore two of the potential explanations for the racial gap in support for BLM: experiences with and feelings toward the police. The experience of being stopped by the police in the last year—whether personally or vicariously through a family member—is associated with less support for BLM, a surprising finding we explore below. The experience of ever having been arrested, by contrast, is not associated with feelings about BLM.Footnote 15 Positive feelings toward the police are significantly and negatively associated with support for BLM. This effect is relatively sizeable: for every standard deviation (SD) increase in support for the police (rating the police about 24 points higher on the 100-point scale), respondents rated the movement about six points less favorably. Collectively, however, these factors only explain a small portion of the gap between white and Black respondents, who remained more than 33 points apart in their ratings even after controlling for experiences with and support for the police.
The finding that personal or vicarious stops by the police were associated with less support for BLM is surprising and deserves more consideration. We know that the quality of experiences with the police is highly dependent on race: Black Americans are significantly more likely to experience mistreatment during stops than are whites (e.g., Reference Epp, Maynard-Moody and Haider-MarkelEpp et al., 2014). Thus, the effect of being stopped may depend on race. Model 1 of Table 4 suggests this is indeed the case, revealing a significant interaction between race and police stops. As illustrated in Figure 1, among Black respondents, the personal or vicarious experience of a police stop was associated with a small increase in support for BLM. For white respondents, by contrast, a similar experience was associated with less support—specifically ranking the movement about eight points more coldly. The most basic issue raised by BLM involves racial bias in police misbehavior. Given that white citizens experience less mistreatment at the hands of the police, it may be that the experience of a more respectful police stop leads white Americans to question the claim of police misbehavior.Footnote 16 Similarly, even among those white respondents who did feel mistreated in a police stop, that experience may lead them to question the claim of racial bias.
Table 4. Coefficients from interactions predicting support for Black Lives Matter
Note: N = 3252.
***p < 0.001, **p < 0.01, *p < 0.05.
Figure 1. Predicted warmth toward Black Lives Matter from interaction of race with police stops
We also consider whether the effect of feelings toward the police depended on race. Model 1 of Table 5 does reveal a significant interaction between race and views of the police. As Figure 2 illustrates, support for the police was only associated with a decrease in warmth toward BLM among white respondents. Despite a wide diversity of feelings about the police, support for BLM remained consistently high among Black respondents. This challenges a popular narrative that support for BLM among Black Americans is primarily motivated by anti-police sentiment. Rather, it suggests that feelings about the police matter—at least superficially—among white respondents.
Table 5. Indirect effects from structural equation model
Note: RMSEA: 0.08; SRMR = 0.05; χ 2 = 1045, df = 45, p < 0.001. Note all mediators standardized.
Abbreviations: BCI: bootstrapped confidence interval; Contrast: indirect effect is significantly larger than other indirect effects; Est.: estimate of indirect effect (Black * mediator).
Figure 2. Predicted warmth toward Black Lives Matter from interaction of race with support for the police
Role of racial animus and resentment
Returning to Table 3, Model 3 adds the other possible explanations: that feelings about BLM may be a product of feelings about Black people or concerns about the relative status of white and Black Americans. Feelings about Black people do appear related to support for BLM: the colder a respondent felt toward “Blacks,” the lower they rated the movement.Footnote 17 This effect is sizeable, translating to about a six-point decrease in support for every one SD increase in coldness toward “Blacks” (rating them roughly 23 points more coldly). Racial resentment was very strongly and negatively associated with favorable views of BLM—the standardized coefficient of −0.48 is easily the largest in the model. A one SD increase in racial resentment was associated with a roughly 16-point decrease in support for the movement. At the extremes, the least resentful were predicted to rate BLM as a 79 while the most resentful were predicted to rate the movement a 23 of 100. Adding these racial feelings and views explains a sizeable portion of the racial gap: after controlling for feelings toward “Blacks” and racial resentment, Black and white respondents were roughly 18 points apart. Put simply, white racism toward Black Americans is an important part of the explanation for why white Americans dislike the BLM movement.
Notably, accounting for racial animus and racial resentment also appears to explain much of the effect of feelings about the police, which correlations reveal is more likely among those high in racial animus and resentment. In other words, “support for the police” appears to be acting in part as a proxy for racial resentment and animus toward Black Americans.Footnote 18
Role of politics
Finally, Model 4 of Table 3 adds measures of political ideology and partisanship, both to explore the role of partisan identification in views of BLM and to see if our proposed explanations—involving the police and race—remain predictive even when politics are included. Those identifying as politically conservative and as a Republican were less likely to feel warmly toward BLM—not surprising given the racialized and partisan use of BLM as an issue in the 2016 election (e.g., Reference Drakulich, Hagan, Johnson and WozniakDrakulich et al., 2017, Reference Drakulich, Wozniak, Hagan and Johnson2020; Reference Updegrove, Cooper, Orrick and PiqueroUpdegrove et al., 2018). In particular, identification as Republican was strongly associated with reduced affection for BLM. The inclusion of political identification reduced the racial gap in feelings toward BLM to about 16 points. The difference, however, remained significant, suggesting the model explains some but not all of why white respondents had less affection for BLM than did Black respondents. Controlling for partisanship also reduced the small remaining effect of support for the police to a role statistically indistinguishable from zero. Notably, racial animus and racial resentment remain strongly predictive of opposition to BLM, with racial resentment retaining by far the strongest standardized coefficient in the model.
Just as we considered race differences in the effects of experiences with and affinity toward the police, we also consider race differences in the effects of feelings and views of race (Model 2 of Table 4).Footnote 19 Although the effect of feelings toward “Blacks” did not appear to differ by race, there was a large significant interaction between race and racial resentment. As Figure 3 illustrates, the large negative association between racial resentment and support for BLM existed largely among white respondents. At low levels of racial resentment, white and Black respondents were similarly supportive of BLM, but support dropped precipitously among white respondents as racial resentment increased. This may not be surprising—especially given the sizeable interracial differences in racial resentment indicated in Table 2—but it does illustrate the powerful relationship between racial resentment and views of Back Lives Matter among white Americans.
Figure 3. Predicted warmth toward Black Lives Matter from interaction of race with racial resentment
Thus, among the four potential explanations for the racial gap in feelings toward BLM, a lack of warmth toward and in particular a resentment of Black Americans among white Americans appeared the most important. The experience of a police stop unexpectedly appeared to decrease support for the movement—though only among white rather than Black respondents—while proclaiming support for the police appeared to largely be a proxy for racial animosity toward Black Americans.
Exploring mediation through indirect effects
The sizeable drop in the coefficient for Black respondents across the models in Table 3 suggests a mediating role for some of our proposed explanations, especially racial feelings and resentments. We explore this mediation more directly through a structural equation model. Table 5 presents just the indirect effects from this model, as well as bootstrapped confidence intervals for these estimates.Footnote 20 We also estimated the contrasts (differences) between these indirect effects. The total effect confirms the racial difference reported above: Black respondents felt roughly 39° more warmly toward BLM. Of the mediators, racial resentment had the largest role, explaining about 10 points of this difference.Footnote 21 Feeling coldly toward Black Americans and identification as a Republican also had sizeable roles mediating the effect of race. Identification as politically conservative had a small but significant role. The indirect effect of race through the experience of a police stop is not significant. As Table 3 revealed, Black and white respondents remain about 16 points apart even after all the mediators are included. In sum, these results confirm the mediation suggested by a comparison of models in Table 3. Specifically, differences between Black and white respondents in their politics, in their animus toward Black Americans, and in their resentment of Black Americans explained a substantial portion—though not all—of the interracial differences in views of BLM. Of these factors, racial resentment played a significantly larger role than the other factors.
DISCUSSION
Substantial racial differences exist in feelings toward the racial civil rights movement BLM. By our estimate from the fall of 2016, white Americans rated the movement about 40 points lower than Black Americans on a 100-point scale. Specifically, the average white respondent only rated the movement a 43 out of 100, meaning they felt coldly toward the movement.
Our central question was why white Americans expressed so much less support than Black Americans. We explored four possible explanations and draw several conclusions. First, experiences with the police mattered, in particular whether the police have recently stopped and questioned you or a family member. However, the effect of these stops depended on race. Among Black respondents, affective support for BLM was slightly higher among those who had recent personal or vicarious experiences with the police but remained relatively high among those who had not. Interestingly, among white respondents, police stops had the opposite effect: they were associated with less support for BLM. In this case, white respondents may have given their personal experiences more weight than descriptions of police behavior seen on the news or social media—those who had relatively positive experiences with the stops may dismiss complaints of police misbehavior, while those with negative experiences may dismiss complaints of police bias.
Affective support for the police appeared to matter more than experiences with the police. However, contrary to the narrative that Black support for BLM is motivated by a dislike of the police, support for the social movement was strong even among the substantial number of Black respondents who felt positively toward the police. Instead, support for the police was primarily associated with less support for BLM among white respondents. Further, the explanation for this race-specific effect appeared to be that “support for the police” was just a proxy for anti-Black sentiment—the association shrank substantially after controlling for racial attitudes.
Of the factors that we considered, the most important explanations for feelings about BLM were racial animus toward and resentment of Black Americans. Racial resentment, in particular, had a very large and negative association with affective support for BLM—specifically among white respondents—and helped explain a sizable portion of the differences in views between white and Black Americans. This association makes sense: racial resentment captures a dismissal of the legitimacy of Black grievances—those rooted in a history of racial exploitation as well as continued discrimination—and instead focuses on Black failings as the root causes of inequalities. Thus, the primary motivation for white opposition to BLM was not support for the police but was instead an animus toward Black Americans and commitment to a racial logic that justifies white privilege.
Views of BLM in historical context
Throughout the long history of the subjugation and exploitation of Black Americans, movements advocating for greater racial equality or justice have been viewed unfavorably by many white Americans, particularly those who were concerned about the relative status of racial groups. This was true for the slavery abolition movement, for the civil rights policies of the Reconstruction Era, and for proposed reforms benefiting Black communities during the Progressive Era (Reference KendiKendi, 2016; Reference MuhammadMuhammad, 2019). Similarly, surveys show low levels of support for the civil rights movement of the 1950s and 1960s among white Americans (Reference LegumLegum, 2017; Reference Sears and KinderSears & Kinder, 1971; Reference Sears and McConahaySears & McConahay, 1973).
The parallels between the Civil Rights and BLM eras are particularly striking and draw attention to some of the forces that may explain similar responses by some white Americans to the civil rights movements in each of the two eras. First, each era was witness to important successes for Black Americans. The 1960s saw the passage of three federal civil rights acts, while the 2008 election of Barack Obama broke a symbolically important racial barrier and served as part of the inspiration for the changes advocated by BLM (Reference TaylorTaylor, 2016). Second, both eras saw dozens of mass protests in predominantly Black urban spaces in cities across the country fueled by broader concerns about racial inequalities and injustices and frequently sparked by specific incidents in which police officers were seen as mistreating or killing Black Americans—including Marquette Frye in Watts in 1965 and Michael Brown in Ferguson in 2014 (Reference Gately and StolbergGately & Stolberg, 2015; National Advisory Commission, 1968; Reference Thorsen and GiegerichThorsen & Giegerich, 2014). Finally, in both eras some politicians explicitly criticized the protesters as disorderly and dangerous, looking to galvanize support among those white Americans concerned about the changes to the racial status quo posed by the movements (Reference Beckett and SassonBeckett & Sasson, 2004; Reference Drakulich, Wozniak, Hagan and JohnsonDrakulich et al., 2020; Reference Drakulich and KirkDrakulich & Kirk, 2016; Reference HochschildHochschild, 2016; Reference TonryTonry, 2011).
There are also important differences between the two eras—BLM emerged amidst a historical backdrop of apparent contradictions on racial progress. On the one hand, real successes in the 1960s ended legal segregation and discrimination and began a long decline in the public's embrace of overt racism as well as a striking increase in agreement on the principle of racial equality (e.g., Reference Krysan and MobergKrysan & Moberg, 2016; Reference Schuman, Steeh, Bobo and KrysanSchuman et al., 1997). Obama's election, the dominance of hip hop in the music industry and in popular culture (Reference ChangChang, 2005), as well as the symbolic successes of numerous high-profile Black Americans, gave rise to talk of a possible post-racial era. More recently, a growing number of white Americans appear to support some criminal justice reforms (e.g., Reference Drakulich and KirkDrakulich & Kirk, 2016; Reference Thielo, Cullen, Cohen and ChouhyThielo et al., 2016; Reference WozniakWozniak, 2020), and formerly radical ideas for achieving racial change appear somewhat closer to mainstream progressive positions—for example, a substantial majority of Democratic candidates for president in 2020 supported at least studying reparations for slavery (Reference KingKing, 2019). On the other hand, actual racial inequalities have remained stubbornly large during this same time, creating a phenomenon Reference TaylorTaylor (2016) describes as “two black societies, separate and unequal” (p. 6). New ostensibly race-neutral policies and practices created since legal segregation and discrimination ended in the 1960s achieve similar ends (e.g., Reference RothsteinRothstein, 2017), and the massive expansion of the criminal justice system preserved racial separation and control (Reference AlexanderAlexander, 2010; Reference TonryTonry, 2011; Reference Wacquant, Blomberg and CohenWacquant, 2003). Further, recent increases in hate crimes (e.g., United States Department of Justice, 2019) and the increased visibility of an overtly racist alt-right movement seem to suggest the resurgence of older expressions of racism (Reference HeimHeim, 2017).
These differences between the earlier and modern eras suggest two distinct stories. First, there have been real successes and there are real opportunities for more progress toward racial equality and justice. Despite visible signs of overt racism, white Americans appear less implicitly or explicitly pro-white since the advent of the BLM movement (Reference Sawyer and GampaSawyer & Gampa, 2018) and may be more open to additional reforms. Our finding of general support for BLM among those white Americans who were low in racial resentment is a reflection of this potential opportunity. However, there are also new barriers facing movements seeking racial equality or justice. The historical transition from more overt to more hidden forms of racist expressions and policies may ironically make them harder to address (e.g., Reference DrakulichDrakulich, 2015a, Reference Drakulich2015b), especially as growing numbers of white Americans now believe that they face more contemporary discrimination than Black Americans (Hochschild 2016; Reference Norton and SommersNorton & Sommers, 2011). Finally, surges in hate crimes and overt racism continue to pose direct threats to Black lives. These changes are only superficially contradictory. As Reference KendiKendi (2016) notes, “racist progress has consistently followed racial progress” (p. xi), and throughout our history surges in antiracism, assimilationist racism, and segregationist racism generally overlap rather than oscillate clearly from one to another. These differences in—and different sources of—support or opposition to the movement reflect these overlapping forces.
Both of these stories are reflected in events in 2020. In the spring, the murder of Breonna Taylor, the release of video from the murder of Ahmaud Arbery, the release of a report showing substantially higher COVID-19 death rates for Black Americans, and the murder of George Floyd all combined to increase support for BLM among white Americans, though a Civiqs daily tracking poll suggests this was short-lived. As the protests and clashes with the police grew in size, scope, and intensity, opposition to BLM shot up among white Americans, leaving only a short period from the middle of May through the end of June when more white Americans expressed support rather than opposition to the movement.Footnote 22 The key finding of this paper—that opposition to the movement among white Americans appears driven by concerns about threats to the racial order—may help explain these patterns as the national story moved from expressions of outrage and sympathy over specific incidents to more explicit calls for real change in broad policies and practices.
CONCLUSION
Our findings have a variety of important implications. The broadest is a reminder to be wary of false narratives masking true motivations, especially for public issues with racial dimensions. The paradox of modern race relations—the persistence of racial inequalities despite improved explicit racial attitudes—can be explained in part by a new racial logic promoting color-blind narratives. Prior work, for instance, has suggested that a focus on individualism to explain opposition to welfare policies or support for punitive policies that disproportionately harm people of color is in fact a cover for explicit and implicit racial animus (e.g., Reference DrakulichDrakulich, 2015a, Reference Drakulich2015b). Similarly, in the present work, we find that declarations of support for the police as a reason to oppose BLM appeared to be a mask for racial animus and resentment. This suggests that statements like “Blue Lives Matter” may be motivated more by an animus toward Black lives than a concern about police lives.
These findings also have more specific implications for several fields of research. Research on perceptions of the police, for instance, often starts from the assumption that such perceptions are primarily about the police—that differences in perceptions must be a product of differences in experiences and interactions with law enforcement. Our results, however, revealed a more modest role for experiences with the police—at least in terms of the overall likelihood of contact—and found that views of a social movement concerned with the police were primarily shaped by views of race. For research on social movements, these findings help confirm the connection between BLM and earlier civil rights movements (e.g., Reference TaylorTaylor, 2016) and shed new light on the ways that opposition to such movements is falsely framed in nonracial terms. They also suggest, as discussed above, that we should continue to expect opposition from white Americans at moments when the possibility of real change to the racial order appears most likely. Research on racism often focuses on views of racial economic inequalities. The present findings—in particular the very strong association between racial resentment and opposition to BLM—reinforce the need for scholars to further explore other dimensions of racial inequality, including inequalities in experiences with the police and the criminal justice system.
Finally, it is important to remember the deeply troubling accusations at the core of the BLM movement: that the very institution and individuals publicly tasked with protecting civilians from harm are in fact causing harm to members of marginalized groups. In a democratic society that values freedom and equality—one which emphasizes basic rights to due process and equal protection—these accusations must be taken seriously as posing a fundamental threat to that democracy. In this light, it is striking that most members of the dominant racial group feel negatively toward a social movement raising these issues. How should we understand this indifference or hostility? Our results suggest it is not primarily out of a lack of experience, nor is motivated by support for the police. Instead, it is a dislike of Black Americans, and a resentment of the very acts of pointing out injustice and asking for equal treatment.
