Why someone did not stop them? Aversive racism and the responsibility of bystanders

The 2020 murder of George Floyd, an unarmed African American man, was committed by a Minneapolis police officer who choked him to death by putting his knee on Mr. Floyd's throat for almost nine minutes. As several other officers watched and did nothing to prevent his death, this incident has provided a powerful catalyst to the persist calls to end racial violence and disparities for African Americans at the hands of police. National protests sparked outcries for police reform, acknowledgment of historic and systemic racism and accountability by leaders to enact meaningful change. Within many debates, analyses and reviews of Mr. Floyd's murder, one question was repeated echoed, “Why didn't the other officers or onlookers stop Mr. Floyd from being murdered”? Some also asked why others stood by looking at the incident, but as Mr. Floyd struggled to breathe, made no other attempts to intercede during this violent situation.

These questions have been asked in many other situations of violence where bystanders were present but did not act or intercede. While cries for justice and indictments for the other officers involved have now taken place, the persistence of two critical questions deserved closer examination. In any critical situation of racial violence, aggression or discrimination, what factors determine or impact the likelihood that help will or will not be provided to a victim? How can we influence behavioral responses of bystanders and equip them to provide proactive intervention in situations of racial bias, discrimination and aggression to prevent dangerous escalation from taking place? While racism exists as a multidimensional and multilevel construct, the focus here is at the individual decision-making and interpersonal levels of analysis as we examine the bystander effect. This is not to negate the importance of institutional or systemic racism, but to focus the current discussion on target–offender–bystander dynamics. This paper uses the aversive racism framework to understand the perception, influencers and conditions under which bystanders fail to respond in situations of racial bias, discrimination and aggression.

The concept of aversive racism is based on the understanding that racial biases are deeply rooted in the US history and reinforced by ongoing societal ideologies (Dovidio et al., 2017). These deeply rooted biases, while powerful, are often expressed without conscious intention or rational awareness by the socially dominant group. While unconscious, these subtle forms of discrimination are coupled with systemic disparities in power and access to resources that have a profound impact on the health and well-being of racialized groups while also serving to reinforce social hierarchy and persistent discrimination (Dovidio et al., 2002). The aversive racism framework acknowledges notions of unconscious or implicit bias and microaggressions but does not ignore the historical context of racial differences that produce actual social, economic and power differences. Using this framework, research on the bystander effect and behavioral intervention models are reviewed to outline a strategy for future research and effective programmatic intervention. This framework may also help to explain why individuals may fail to intervene as bystanders in situations of racial bias or discrimination, yet maintain a self-perception of being egalitarian, non-racist and unbiased.

The recent and high-profile incidents of racial violence in the USA (e.g. George Floyd, Breonna Taylor, Eric Garner, Stephon Clark, Philando Castile, Michael Brown, Trayvon Martin, Tamir Rice, Walter Scott and others) have sparked national and global protests focusing our attention on all forms of racism, especially involving race-related police violence (Gray and Parker, 2020; Rivera and Ward, 2017). However, disagreements over issues such as the use of “black lives matters” versus “all lives matter” to the debate over whether systemic racism exists within law enforcement require the critical theoretical lens of aversive racism to better understand and address. As Konrad (2018) argues that, “distancing from racism has allowed contemporary American extremists to reframe themselves as victims of closed-minded progressives seeking to elevate undeserving and/or dangerous out-groups at the in-group's expense. Effective anti-racism techniques are needed to counter implicit biases to limit the attractiveness of extremist views” (p. 14).

Thus, this paper makes the case that the aversive racism framework calls attention to persistent anti-racial sentiments among seemingly well-intentioned individuals within the context of differential access to power, privilege and resources that produce significant disparities in the experiences, interactions with and lives of African Americans. As a result, recognizing unconscious bias is necessary, but insufficient. Thus, I argue that solutions to persistent racial inequities must go beyond mere awareness or consciousness raising efforts. The impact of persistent racial bias, discrimination and racial violence is not only committed by perpetrators but is also facilitated by otherwise well-intentioned individuals who fail to act or intercede when racial incidents or acts of violence occur. Thus, in additional to raising awareness, the aversive racism framework points to the need for facilitated behavioral changes not only for the perpetrators of racial violence and hatred but also for well-intentioned bystanders who fail to act or intercede in these situations. Examining aversive racism and the existing research on effective bystander intervention training highlights the need and importance of shared responsibility in preventing the outcomes of racial violence and discrimination as we seek to create meaningful and long-lasting social change.

The origin of aversive racism began with attempts to explain the discrepancy between the endorsement of egalitarian values by white Americans and acts of racial bias that prompted the need to draw attention to racial bias and discriminatory acts that were being committed by otherwise “good-intentioned” individuals (Dovidio et al., 1997). This early work asked the questions of how well-intentioned individuals can commit and/or support both blatant and subtle forms of racial bias, discrimination and violence, yet not recognize this incongruity. The disconnect between expressed egalitarian attitudes versus racially biased actions is not only perplexing but also can service as a source of what is often referred to as “racialized identity trauma” for targets of racial bias and violence (McCluney et al., 2017). This means that, in addition to the experience of those who are the direct targets of racial discrimination or violence, racial trauma (e.g. witnessing the killing of an African American by white police officers) produces psychological injury that also impacts those who witness and share the same racial category or identity group. Whether experienced directly or through observations (e.g. news reports, social media, etc.), this racialized trauma can overwhelm a person's capacity to cope, cause harm, anxiety and threaten overall psychological well-being (Kira, 2010; Bryant-Davis, 2007; Sewell and Jefferson, 2016). Thus, seemingly neutral interactions with well-intentioned individuals who believe themselves to be unbiased can actually have a harmful impact on psychological well-being and produce a sense of “threat” among those within the same racial group as the victim (Emerson and Murphy, 2014).

This racialized trauma can be triggered by those who possess unconscious forms of racial bias but engage in the rationalization of discriminatory actions because they see themselves as fair and egalitarian (Pearson et al., 2009). In fact, according to the aversive racism framework, individuals who commit discriminatory acts may also sympathize with victims of past injustice, genuinely support principles of social equality and regard themselves as unbiased, fair and non-prejudiced individuals. This is clearly illustrated in the stark racial differences in beliefs about racial bias in policing (Leopold and Bell, 2017), despite data showing that African Americans are twice as likely to have force used (or threatened force) against them by police (Vito et al., 2020).

Since the initial development of the aversive racism framework, there has been substantial research and applications over the past decades (Dovidio et al., 2017). The aversive racism framework has been applied to decisions concerning qualifications of job applicants who are recommended for hiring (Jones et al., 2016), applicant recommendations for college admission (Hodson et al., 2002), the impact of inadmissible DNA evidence on judgments of guilt or severity of sentencing of black versus white defendants (Hodson et al., 2005), employment decisions (Dovidio and Gaertner, 2000) and medical interactions between black patients and white physicians (Penner et al., 2014; Penner and Dovidio, 2016). Within this research are robust findings showing that across time, context, population, and outcomes, the aversive racism framework is a powerful explanatory tool for understanding bias among people who may be well-intentioned, but, nonetheless, engage in discriminatory actions and decisions. While unintentional, aversive racism has been consistently shown to produce disparate outcomes that can be as negative and debilitating for the target minority groups as intentional or overt forms of bias (Dovidio et al., 2002). Furthermore, because of this subtle bias, efforts to detect and eliminate aversive racism must be different from efforts to address more direct or overt forms of racial bias and discrimination.

Thus, the aversive racism perspective calls into question methods for addressing bias and discrimination, especially when more subtle forms of bias tend to be ignored (Abelson and Ettlin, 2004). For example, Murrell and colleagues (Murrell et al., 1994) studied attitudes toward affirmative action as a function of targeted group, framing of the policy and institutional context. Resistance to affirmative action was aroused more by policies specifying African Americans as the targeted group and by policies presented without justification. Supportive of the aversive racism framework, the level of resistance to the policies presented without justification for African Americans as the target group was higher than for all other targeted groups with or without justification.

Rodenborg and Boisen (2013) connected the aversive racism framework and intergroup contact theory to make a case for lack of cultural competence and persistent racial segregation. Sue (2010) observed that, “racial microaggressions are most similar to aversive racism in that they generally occur below the level of awareness of well-intentioned people” (p. 9). Thus, efforts toward anti-bias education programs (Kalev et al., 2006; Yeager and Walton, 2011) often prove to be ineffective in producing the intended outcomes, as predicted by the aversive racism framework. As Dovidio and Gaertner state, “like a virus that has mutated, racism has also evolved into different forms that are more difficult not only to recognize but also to combat” (Dovidio and Gaertner, 1998, p. 25).

The irony is that aversive racists profess egalitarian attitudes, yet engage in subtle forms of discriminatory behaviors and avoid overt displays of bias in intergroup exchanges. However, because many aversive racists are unaware of their own inconsistencies, their biases still produce discriminatory behaviors that they may be unaware of, or that they do not acknowledge as discriminatory in nature. These behaviors can include nonverbal signals (e.g. physical distancing) as well as verbal comments (e.g. all lives matter) that may not be recognized as biased by well-intentioned individuals but may be experienced as demeaning or threatening by racial group members (Sue, 2010).

For example, Dovidio et al. (1997) found that whites' racial bias predicted negative nonverbal behaviors, reflecting discomfort (rate of blinking) and dislike (gaze aversion). These socially distant and dismissive behaviors can be experienced by African Americans as invalidations, microaggressions or targeted acts of discrimination (Sue, 2010). In other words, the differential experiences or interpretations of behavior have clear impact on the physical, social and psychological well-being of individuals is unintentional, but, nonetheless, produces negative experiences and outcomes (Williams and Clarke, 2019).

To preserve self-perceptions of fairness and equality, these consequences are often blamed on the individual and discounted as having roots in systemic bias and discrimination. For example, Penner et al. (2017) show that relative to medical interactions between a doctor and patient of the same race, racially different medical interactions are shorter in length, involve less positive affect and are less patient-centered. While explicit procedures may be followed, subtle differences remain. White physicians spend significantly less time answering questions, providing health education, planning treatment, building a relationship with African American relative to white patients and make less effort to involve African American patients in medical decision-making. However, white physicians' explicit racial bias has been shown to be a weak predictor of the overall quality of patient care outcomes (Hagiwara et al., 2016; Hagiwara et al., 2017). Subtle forms of aversive racism are masked by explicit rejection of overt racial bias. Other research shows similar effects within educational, business, public and other settings (Aberson and Ettlin, 2004; Apfelbaum et al., 2008; Holmes, 2010; Nail et al., 2003; Toosi et al., 2011). Thus, the aversive racism framework can explain persistent racial disparities, differential perceptions and experiences of well-intentioned majority group members and enduring psychological trauma and consequences inflicted upon racialized individuals (Dovidio et al., 2016).

Research also shows that the aversive racism paradigm actually explains some of the underlying reasons for why race-specific experiences and interactions persist in violence experienced by African Americans at the hands of the law enforcement personnel (Trochmann and Gover, 2016). Thus, when confronted by the possibility that their behavior is racially motivated, aversive racists actively seek nonracial explanations for their behavior to preserve a non-prejudiced self-image (Dovidio et al., 2017). So, what does this mean for programmatic efforts to address the frequency and impact of racial bias and discrimination that persist?

Many argue that programs that work at the individual level, focus on misperceptions that are common to everyone (e.g. we all have unconscious biases), are not effective (Noon, 2018). Others argue efforts focused on raising awareness can elicit defensiveness or produce backlash that may increase rather than reduce racial bias and discrimination (Legault et al., 2011). Some contend that a focus on unconscious or implicit bias can produce accusations that these programmatic efforts are subtle forms of anti-white attitudes and constitute reverse discrimination (Norton and Sommers, 2011). However, examining why individuals choose not to act or intervene in situations related to racial bias, microaggressions or violence can be explained by the broader lens of aversive racism. Using this framework explains bystander behavior and points to both individual and contextual factors that influence bystander behaviors as well as behavioral methods that can influence outcomes in a proactive rather than evaluative manner.

Gaining a better understanding of the perceptions, decisions and behaviors of bystanders in crisis or emergency situations has a substantial amount of existing research. Originally called the “bystander effect,” research shows that an individual's likelihood of receiving help actually decreases when others are present (Latane and Darley, 1970; Latane and Nida, 1981). Within this extensive body of work, both the context of the situation and aspects of bystanders influence whether to or not to help is attempted or provided. Factors such as type of situation, location of the research, characteristics of the victim, attributes of other bystanders, ability of bystanders to communicate with one other and the level of expected danger have been shown to consistently influence individual responses as well as outcomes (Fischer et al., 2011; Latane and Nida, 1981). In addition to competence, demographic characteristics of bystanders were identified in early research by Gaertner et al. (1982), showing that blacks received less help from white bystanders. The bystander effect was more pronounced for those high in prejudice attitudes, especially among white bystanders.

Also relevant to the current context are the findings that the competence of bystanders and the race of the victim have been shown to impact responses and outcomes. For example, bystanders who are characterized as high competence increased the likelihood that help is given to victims (Horowitz, 1971; Smith et al., 1972). Bystander competence is defined as the awareness to notice and perceive an event (or warning signs), define the event as requiring action/intervention, taking responsibility for acting (i.e. feel a sense of personal duty) and having a sufficient level of self-efficacy (i.e. perceived competence to successfully intervene) to effectively intercede (Cramer et al., 1988; Shotland and Heinold, 1985; Pantin and Carver, 1982). Self-efficacy is important here because it can clearly be enhanced or developed through experience and training (Kettrey and Marx, 2020). Ironically, the presence of other officers who failed to intercede in the George Floyd murder could have provided a strong signal to others that action was not warranted or effective because some may have judged these officers as “competent responders” (Franzen, 1999; Horowitz, 1971; Krueger and Massey, 2009). This suggests that the ability and actions of bystanders to intercede in situations involving racial aggression and violence can be influenced or changed.

Other research on the bystander effect focused on key psychological processes that have a significant impact on bystander responses (Fischer et al., 2011). Diffusion of responsibility is often noted as having a significant impact on the tendency of observers to cognitively share or divide the responsibility by the number of bystanders that are present. The more that others are present, the less “responsibility” is perceived by an individual bystander. Also, relevant is the evaluation apprehension process or the fear of being judged by others in situations where there are social or relational ties between bystanders (e.g. others same racial identity group members, friends, coworkers). The concept of pluralistic ignorance has also been identified to influence bystander responses defined as the tendency to rely on the overt reactions of others to shape how one defines and responds in ambiguous situations (Fischer et al., 2011). This notion of pluralistic ignorance is similar to concepts such as conformity or group cohesion pressures known to have a strong influence on the perceptions and behavioral responses of others (Jones et al., 2016). These factors have been called “process ambiguity,” which can shape or distort how situations are defined in terms of how appropriate a response is judged to be among bystanders (Fisher et al., 2011). One could argue that aversive racism may produce racially biased process ambiguity, thus increasing the bystander effect when the victim is a person of color.

A meta-analysis of the bystander effect (Fischer et al., 2011) examined a range of different factors among previous studies that measured either actual helping response or response latency. Their findings clearly show that the bystander effect is reduced when the situation or emergency is perceived as dangerous, when the perpetrator is physically present or when bystanders perceive a cost to intervene. This effect is also reduced when a majority of bystanders were male (or those with physical strength) or when friends and acquaintances are present. Additional bystanders who are perceived as sources of social, physical and psychological support can increase the likelihood that victims receive help, thus reducing the bystander effect. These findings are consistent with a traditional “cost–reward model” (Dovidio et al., 2016), arguing that heightened arousal sources (e.g. perceived benefits or consequences) are a necessary condition for bystanders to act versus internal affective conditions such as sympathy or empathy alone. However, judgments of cost-versus-benefits of helping are clearly impacted by aversive racism and racialized unconscious bias. This has been illustrated by empirical work that induces the experience of racial hypocrisy, which increased the perceived costs of no acting/intervening (Son Hing et al., 2002).

The irony of this meta-analysis is that its findings would have predicted greater intervention in the George Floyd murder than what actually took place. Bystanders inhibitions are less pronounced in clearly dangerous situations or when others who share social ties are present. However, none of these traditional findings helps explain the failure of bystanders (especially other police officers) to intervene and prevent the death of this unarmed African American man. While some attributes of bystanders and victims were included in their meta-analysis, race of the victims and racial attitudes of the bystanders were not measured due in large part because of the limited number of previous research studies that examine these critical factors. It could be that race and racial attitudes alter the nature of bystander inhibition and change the conditions under which the “cost–reward–arousal” model would predict whether bystanders choose to intercede. In other words, we must examine other explanations for why otherwise well-intentioned individuals fail to intercede in a situation such as the world observed in the murders of George Floyd and others.

Previous research on the bystander effect assumes that judgments and decisions about whether to intercede and provide the needed help are influenced more by situational factors and victim characteristics than other factors. However, early work on aversive racism and lack of response in helping behavior for African American victims may provide a different explanation. Silence, inaction and failure to speak out in the face of racial injustice, discrimination and race-based violence require a model or theory that speaks more directly to the nature of the bystander than the victim or the situation. Some argue that fear of repercussions, situational ambiguity and misinterpretation of the situation are driving factors. However, these factors also may fall short in explaining how the race of the victim alters the well-documented nature and parameters of the bystander effect. Sue et al. (2019) argue that, “in many cases, bias and discrimination go unchallenged because the behaviors and words are disguised in ways that provide cover for their expression and/or belief that they are harmless and insignificant” (p. 128). This statement suggests a potential linkage between aversive racism and the bystander effect.

Thus, examining issues or bystander action (and inaction) through the lens of aversive racism becomes necessary and essential. This is due in large part to one of the core principles of aversive racism, which is that racial prejudice and bias are expressed in subtle yet rationalized ways to protect egalitarian self-perceptions. These egalitarian self-perceptions can be heightened by professional identities that have strong values-based components such as the ethic of care in medicine (Johnstone and Kanitsaki, 2010) or public safety and justice in policing (Oberweis and Musheno, 1999). As Gaertner and Dovidio (2000) outline, most whites are motivated to avoid personal costs imposed for responding negatively to African Americans in interactions that conflict with their self-perception of being fair and egalitarian. This means that for individuals who are high in aversive racism ideologies, the cost–benefit–arousal conditions are different that their non-aversive racism counterparts. Earlier, Gaertner et al. (1982) stated that, “In fact, most whites may be motivated to avoid a special cost imposed for responding negatively to a black person: the personal attribution of bigoted intent” (p. 70). Thus, nonracial factors are amplified to justify or rationalize responding in a negative manner, especially in emergency situations involving an African American victim. The use of non-race-related explanations for not interceding are often justified by attributions that help was not necessary, warranted or deserved. Factors under the category of “process ambiguity” such as diffusion of responsibility are used to not only justify lack of intervention but to protect the individual's self- perception of racial neutrality and fairness. Similarly, evaluation apprehension and audience inhibition introduce strong normative and self-presentation pressure to appear “non-racist” that would decrease the likelihood that African American victims would receive help, especially when more than one white bystander is present.

In other words, the outcomes of the bystander effect may appear similar on the surface, but the underlying factors and explanations differ according to the aversive racism framework. As a result, strategies to change or reshape behavior must be viewed through the appropriate theoretical lens. This means providing strategies and skills that override the national tendencies to engage in self-protective efforts and rely on non-race-related rationales for lack of action or not providing help. It may be especially important to provide behavioral interventions that override default preferences and increase the likelihood that individuals who perceive themselves as non-racist will not misdiagnose a situation as one that does not need or warrant intervention as a way to protect their own sense of egalitarian and non-racist self-perception.

Using an aversive racism framework necessitates going beyond awareness raising of implicit or unconscious bias because these approaches will likely be ineffective among individuals who perceive themselves as holding egalitarian values with strong self-perceptions of being anti-racist. Some call attention to anti-racist techniques such as “micro-interventions” where there is intentional practice of non-racist behaviors that are similar to training of allies used within-diversity and inclusion programs (Nair et al., 2019). As Sue et al. (2019) state, “everyday interventions of allies and well-intentioned bystanders have a profound positive effect in creating an inclusive and welcoming environment, discouraging negative behavior and reinforcing a norm that values respectful interactions” (p. 132). They argue that providing behavioral change through training can alter organizational and eventually societal norms that will not take place through individual motivation or awareness alone.

Some intervention efforts are already showing positive outcomes. For example, Dovidio and Gaertner (2000) examined the impact of aversive racism on selection decisions and found the use of blind evaluations in both selection and assessment show positive outcomes for reduction discriminatory judgments. Ferdman et al. (2020) make a case for inclusive leadership as a tool for creating broad organizational and societal change. Some argue for the need to challenge current definitions and concepts of key constructs that are used in judgments and decision-making. For example, Ely and Meyerson (2010) directly challenge the traditional notions of leadership in their examination of organizational safety and effectiveness for offshore oil drilling organizations. Other agree that starting programs early in the leadership development cycle that focus on inclusive leadership, and its link to ethics and social responsibility are needed (Murrell et al., 2020).

This paper began by asking the question of why some individuals do not intervene in situations involving racial aggression and raced-based violence as witnessed in the murder of George Floyd. Understanding the interconnection between the aversive racism framework and the factors that produce the bystander effect provide important insight into this timely question. However, what is left to address is how to change the current conditions to increase the likelihood that those who are faced with racial aggression or violence receive the aid and support needed. Some of the extensive work on bystander sexual assault prevention programs may provide some insight into how we should approach similar efforts in the context of race-based assault and violence. This approach is based on the expectation that equipping numerous bystanders with behavioral strategies to intercede will increase the likelihood that racial aggressions and acts of violence may be deescalated before another death occurs. The focus on bystander invention behavioral training does not, however, negate the need to address systematic and institutional racism that manifests as unethical or illegal behavior among some individuals within law enforcement and within the criminal justice system (Trochmann and Gover, 2016). Work on bystander training used in employee development shows that by focusing on bystanders as potential allies, intervention efforts are experienced as less threatening than traditional programs that tend to approach participants as either potential victims or potential perpetrators (Scully and Rowe, 2009). While promising results, more longitudinal research is needed to provide evidence of the long-term effectiveness of these behavioral interventions. However, bystander development efforts may reinforce the shared responsibility of individuals as co-owners of the solutions to address negative or damaging behaviors from microaggressions to overt acts of racial violence (Nelson and Dunn, 2011). These types of interventions may also build societal and organizational capacity to engage in socially responsible actions through a long-term commitment to developing individual competency and mutual accountability (Vasquez et al., 2019). Again, additional work that bridges traditional research and practice on the connection between behavioral interventions and bystander intervention is needed.

Recently, Kettrey and Marx (2020) conducted a review and meta-analysis of bystander sexual assault prevention programs. They examined outcomes associated with various intervention approaches such as developing knowledge of effective strategies, recognizing a sexual assault event, taking responsibility and identifying a situation as appropriate for intervention or actions as specific behaviors to develop engaged bystanders who elect to commit constructive responses to racialized situations. This suggest that any intervention or training program must be structured to go beyond “mere awareness” to be effective. Across these studies, the impact of bystander training that focused on awareness only – that is taking notice of warning signs of sexual assault – was not impactful. Programs aimed at helping individuals identify situations as either appropriate or inappropriate for intervention showed only weak outcomes. However, promising effects were shown when programs were aimed at specific behavioral change such as the actual practice of intervening in situations that may occur following training. This shows promise for bystander intervention programs that are competency-based or behavior-focused in terms of their content and measurable outcomes. Other work has shown positive outcomes when program models are skills-based and focused on bystander intervention behaviors, rather than direct attempts to change attitudes alone (Seen and Forrest, 2016). Clearly, additional research on how effective these various intervention strategies can be to impact individual, institutional and ultimately societal influences on bystander responses to racial microaggressions and violence in needed.

Similarly, Sue et al. (2019) identify a list of intervention strategies that show promise in responses to the de-escalation of microaggressions in a workplace environment. These programs are specifically targeted toward anti-racist behaviors, although not necessarily created based on the aversive racism framework. Their review examined what authors labeled as “micro-intervention strategies” across three categories: those directed toward individual microaggressions, those directed toward institutional macroaggressions and those directed toward societal macroaggressions. Within each of these three categories are a range of specific behavioral goals identified. Goals include making the “invisible” visible by altering the cognitive processing and labels placed on situations involving racial and other forms of aggression.

Within the aversive racism framework, this would focus attention on changing meta-communication used to reinforce non-racist self-perceptions as well as examining structures and systems that perpetuate racial disparities. A second category is aimed at disarming or deconstructing both micro- and macro-aggressive behaviors. Interestingly, this includes focused public actions such as boycotts, strikes and protests, which provide validation to the need for individual and collective voice in the wake of racial and other forms of injustice. The third category points to the need to “educate the offender” and includes consensus-building as part of organizational and societal efforts to transform individuals, systems and structures that create and perpetuate inequities and all types of aggressions. Finally, the fourth category involves seeking external reinforcement or support to provide long-lasting and sustainable solutions such as laws, public policies and organizational practices. Interestingly, this is where aspects of inclusive leadership development and accountability become critical to ensure long-term social change (Murrell et al., 2020).

While this model exists as a relatively recent framework, the possibility of future research and practice to both validate and implement these suggested strategies outlines a promising agenda. Their model also has the advantage of not ignoring the different levels that are necessary to alter processes, impact and outcomes consistent with the aversive racism framework. Focusing on behavioral responses also shifts the attentions away from issues of intent, motive and attitudes, which we know to be unreliable based on the motivation to preserve egalitarian self-perceptions. Targeting behavioral change among a diverse array of bystanders has the promise of building greater capacity than attempts to change the behaviors of perpetrators alone. This model placed within the context of the aversive racism framework also has the benefit of for equipping or “arming bystanders” to move from passive observers to active agents of behavioral and social change (Sue et al., 2019).

However, caution is necessary to avoid placing the burden exclusively on individual behavioral change among well-intentioned but nonetheless biased observers as the singular pathway to preventing racial violence and systemic racism. Equipping individual bystanders is important for creating meaningful change but also will not erase historic, social, economic and systemic racism that also contributed to the death of George Floyd and numerous other victims.

As the aversive racism framework outlines, individual desires to appear and perceive oneself in a positive and non-racist manner are significant obstacles to overcome. This goes beyond “denial” as often referenced in work on unconscious bias and toward acts of racial “distancing” (Konrad, 2018). This includes denying of facts, incidents or concrete examples (“those statistics are wrong”), developing alternate explanations (“only a few bad apples are racist”), reinforcing beliefs about meritocracy and deservingness (“they were likely guilty of something”) or claims of being colorblind (“I don't see race”). While each of these strategies as well as others may retain an individual's sense of egalitarian values and anti-bias self-perceptions, they can also cause of a lack of action or freezing in situations involving racial bias and discrimination (“I just don't know what to say or do”). This lack of action among well-intentioned observers can provide a silent acceptance or validation of racial bias and signal “silent agreement” with racism and its consequences. As seen in reactions to the “black lives matter” and related movements, public statements of condemnation can be judged as insufficient or insincere and backfire against those perceived as having a history of racial bias, discrimination or inequality (Leopold and Bell, 2017; Opie and Morgan-Roberts, 2017; Vasquez et al., 2019). Meaningful invention and social change are necessary but must begin with accountability from those who hold the means of power and privilege, yet choose to remain passive bystanders rather than become active allies. As Rev. Dr. Martin Luther King Jr. said in his 1965 speech discussing civil rights: “In the end, we will remember not the words of our enemies, but the silence of our friends.”