In recent years, murders of BlackFootnote 1 men by White police officers have sparked outrage around the world. In Brazil, in 2013, Amarildo Dias de Souza, a Black man who worked as a bricklayer helper and who lived in Rio de Janeiro’s Rocinha slum, disappeared after a police operation. Twenty-five officers were sued for Amarildo’s disappearance, generating eight convictions, but the case is still under investigation. In 2014, a White officer murdered the 18-year-old Black American Michael Brown. Investigations showed that the young man was unarmed. The event sparked protests across the country (Chaney, 2015). These are just two examples of the many cases where police officers abused their power during operations (Barrett et al., 2014; Cerqueira & Moura, 2013; Sinhoretto, Batitucci, et al., 2014a).

The police force, an armed corporation responsible for maintaining and restoring order, is the primary public security service with which the community interacts. To pursue its mission, it carries out ostensible activities such as conducting stops, which can be defined as the “technique of stopping an individual or a group of people . . . with signs of suspicion; who have practiced or are on the verge of committing criminal offenses, to investigate, guide, warn, arrest, assist, etc.” (Boni, 2006, p. 29). Police stops should limit individual freedom according to the collective interest and be based on founded suspicion—a concrete and sure presumption of the level of suspicion of the individuals stopped (Boni, 2006). Although most police forces have operational manuals to guide their procedures, the way an officer conducts a police stop or search depends on their personal evaluation, which may be influenced by negative attitudes, for example, about the suspect’s race.

An attitude can be seen as a favorable or unfavorable evaluation of a stimulus or set of stimuli (e.g., Eagly & Chaiken, 1993). Attitudes regarding racial issues have been discussed with related concepts such as (a) racism, which can be seen as an ideology based on the belief that biological, phenotypic, or cultural differences are inherently related to moral, intellectual, and aesthetic characteristics of individuals (Munanga, 2004); (b) racial prejudice, a set of negative attitudes regarding an individual or group of individuals, based on the characteristics of the individual/group, such as skin color, type of hair, and size of lips (Nogueira, 2006); and (c) institutional racism, a systemic form of prejudice, which can be defined as “the failure of institutions and organizations to provide a professional and adequate service to people due to their color, culture, racial or ethnic origin” (Geledés – Instituto da Mulher Negra, 2006, p. 22).

The influence of police officers’ racial prejudices and the institutional racism of the police on Black individuals has been found, for instance, in Brazilian studies that interviewed police officers and individuals who were stopped (Ramos & Musumeci, 2004; Souza & Reis, 2014), directly observed police stops (Ramos & Musumeci, 2004; Silva, 2009; Sinhoretto, Batitucci, et al., 2014a), analyzed police lethality data (Cerqueira & Moura, 2013; Sinhoretto, Silvestre, & Schlittler, 2014b), and examined statistical data resulting from the formalization of complaints (Duarte et al., 2014). This is a fundamental social issue, as the alarming number of young Black men killed in police encounters demonstrates. As one example, Sinhoretto, Silvestre, and Schlittler (2014b) analyzed 734 cases of deaths resulting from police conduct in the state of São Paulo, Brazil, between 2009 and 2011 and found that 61% were of Black individuals—whereas only approximately 30% of the population of São Paulo is Black.

This differential treatment of people of color can produce a distrustful relationship between the population and the police. For example, in a Brazilian survey on racial and ethnic characteristics of the population (Instituto Brasileiro de Geografia e Estatística, 2011), 68.3% of the participants considered that justice and police were the fields with the greatest influence of racial biases, suggesting that Brazilians do not trust their own public security system. The consequences of this inefficiency have been discussed in the literature as institutional racism, which can be seen in the treatment of Black users by institutions in different areas, such as education, health, and security.

Of course, racially skewed police stops are not exclusive to Brazil. Since the 1990s, after episodes of police violence directed at young Blacks, the United States has intensified the assessment of police practices in addressing so-called street crimes. For instance, Gelman et al. (2007) analyzed 5,000 cases of Stop, Question, and Frisk (SQF) police approaches (also known as Terry stops in some states), a U.S. street police department policy that allows police officers to search civilians behaving suspiciously, looking for weapons or smuggled items. The results indicated that more Blacks and Hispanics were stopped, even when the study controlled for geographic variables, such as racial characteristics of the neighborhoods and criminal indexes of each district. Thus, Black and other underrepresented individuals are more prone to being stopped by the police. And during stops, they are more prone to suffer unnecessary use of force (Cerqueira & Moura, 2013; Sinhoretto, Batitucci, et al., 2014a). This produces immense behavioral consequences for both their relationship with the police force itself and other areas of their lives. For instance, the number of occurrences that involve underrepresented groups could be higher due to the disproportional number of times they are stopped. In fact, Mizael and Sampaio (2019), in a behavioral and cultural analysis of the Brazilian police force, argued that the “war on drugs” (the policy of trying to reduce drug trafficking at any cost) leads to greater policing of underprivileged areas, where there are more Blacks and people of color. Additionally, the unnecessary use of force toward underrepresented groups could lead to responses such as escape and avoidance in the form of freezing or hiding from the police, which could be seen by police officers as suspicious behavior, leading to more stops (and more fear and escape or avoidance responses by the individuals being stopped). Additionally, the unnecessary use of force could lead to a difference in attitude toward the police: Whereas Whites could be more prone to see the police force as a positive and necessary group in society, Blacks and other underrepresented groups could be more likely to associate the police force with negative attributes.

In this respect, any behavior analyst or social scientist working with this public or interested in them should be aware of the most recent evidence regarding the matter. The first step to planning any type of intervention is being aware of the current data. This facilitates the identification of gaps and the proposition of additional research and future application. More importantly, any professional interested in contributing to alter this fundamental social issue should be familiar with empirical work that could inform organizational and cultural interventions (e.g., training and educational activities, culturo-behavior science and organizational behavior management applications) to pursue that end. In this direction, we carried out a systematic review of the literature to analyze recent results of empirical, archival, and statistical research that investigated the relationship between racial prejudice and police stops.

Method

We searched PsycInfo, which covers themes of the behavioral and social sciences, with keywords related to racial issues and police stops: [race OR racial OR ethnic] AND [“police stop” OR “traffic stop” OR stop AND search OR stop AND frisk]. We also searched for the Portuguese translation of the keywords. Keywords must have appeared in the abstract. The search was filtered by type of document (journal articles only) and time period (2014 to 2019) and was carried out in October 2019. The inclusion criteria for the studies were as follows: The study had to be an empirical research (e.g., reviews and mathematical modeling studies were excluded) and to have primarily investigated the relationship between racial prejudice and police stops. The selected studies were analyzed regarding (a) authors, (b) aims, (c) study design, (d) participants, (e) country, (f) dependent variables, (g) independent variables, and (h) main results found. The recommendations of the PRISMA group (Liberati et al., 2009) were considered for the presentation of the results.

Results

In the identification stage, no studies written in Portuguese and 23 studies written in English were found (with no repeated results). Of these, seven studies were eliminated: two for being nonempirical and five for not being primarily related to racial prejudice and police stops. Thus, 16 studies were read in full and evaluated for eligibility, all of them being included. Table 1 summarizes the reviewed studies’ main characteristics. The studies were carried out in the United States, England, Wales, and the Netherlands and mainly analyzed database or survey data.

Table 1 Summary of the Studies Reviewed
Full size table

The studies’ main results on the relationship between racial prejudice and police stops showed, first, that a larger proportion of Black people (vs. other races), especially men, were stopped and exposed to excessive use of force by police (Ariza, 2014; Morrow et al., 2017; Motley Jr. & Joe, 2018; Svensson & Saharso, 2015). Black men with low incomes were especially prone to this (Motley Jr. & Joe, 2018). Furthermore, living in communities where the use of force by police is common or racially skewed was associated with poor health (Sewell, 2017).

As for the perceptions of interactions with the police, young Black participants reported having difficulty communicating with officers—especially those participants who were not proficient in English or were afraid to ask questions, arguing that the police could see this type of behavior as defiance behavior (i.e., “being smart”; Barrett et al., 2014). They also reported abuses of power and “overpolicing” directed toward them (i.e., a more intense and frequent policing than necessary to maintain order; Barrett et al., 2014). Additionally, research has found that Black participants reported more negative attitudes toward the police (Nadal et al., 2017). The only experiment included in our review further found that social contextual framing factors, such as the local climate of police–community relations, the legality of the stop, and the officer’s work history, have a similar or even greater effect than the race of participants on how citizens perceive and evaluate videos of actual police–citizen encounters (Braga et al., 2014).

During traffic stops, Black drivers were also stopped more often during the day than Whites, suggesting a racial bias, because it is harder to ascertain an individual’s race in darkness (Ritter, 2017; Taniguchi et al., 2017). In addition, specific police units (e.g., the High risk Entry and Arrest Team, HEAT, team) presented greater racial disparity in the number of stops conducted (Taniguchi et al., 2017). Blacks and Latinos received more traffic tickets than White drivers, and Black individuals were disproportionately targeted for discretionary searches (Gilliard-Matthews, 2017). Compared to White drivers, African American drivers were more likely to experience an investigative stop, to experience more invasive inquiries in this kind of stop, and to receive a formal complaint (Epp et al., 2017).

The study by Vito et al. (2017) found that traffic stops with Black drivers were less likely to generate citations compared to that of other races, but the researchers suggested that this could be explained because officers believed Blacks were involved in more serious crimes. According to Voigt et al. (2017), police officers’ utterances spoken to White community members during vehicle stops were consistently more respectful, but not more formal, than those directed toward Black individuals, even after controlling for contextual factors such as the severity of the offense or the outcome of the stop. Related to that, traffic stops conducted at deteriorated settings were associated with a driver’s race being described as African American (Barnum & Miller, 2014).

One study even found that officers’ categorizations of Black women’s body weight in police stops’ reports were racially biased, with Black women being more likely than White women to be labeled heavy and to be stopped in private rather than public spaces, even after controlling for diverse variables (Kwate & Threadcraft, 2015).

Discussion

Institutional racism can be measured by the low level of access, quality, or satisfaction with products and services (e.g., education, health) provided to specific racial/ethnic group(s) by an institution. This form of racial prejudice is a fundamental element in the maintenance of racial inequality in a society. As stated in the Brazilian Handbook on Fighting Institutional Racism, “What institutional racism produces is not only the lack of access and the low quality of access to services and rights, but it is also the perpetuation of a structural condition of inequality in our society” ( Geledés – Instituto da Mulher Negra, 2006, p. 13).

Taking those measures of institutional racism into account, the studies we reviewed clearly demonstrate the institutional racism present in police stops—at least in the countries where the research was conducted. Black people are systematically harmed by police actions, as evidenced in the studies reviewed. Having a biased police force interferes with the population’s trust in the alleged aims of the police—that is, safety and order—especially that of people of color and other underrepresented individuals. That produces a negative perception of police officers, reported by participants from studies carried out in three countries: the United States (Braga et al., 2014), England (Barrett et al., 2014), and the Netherlands (Svensson & Saharso, 2015).

The century-old idea that Black and other non-White individuals are inferior influences the behavior of members of a society, being reinforced in and by several institutions (e.g., religious, educational). Being present and reinforced in several contexts, this incorrect and outrageous idea becomes naturalized—that is, people tend to see it as a normal or natural feature of the environment, and they tend to avoid any form of countercontrol (see Sidman, 1989). However, a system that maintains institutional racism directly interferes with the quality of life of those targeted by racial selectivity during stops and can generate popular protests and riots. In Nadal et al. (2017), for example, a 19-year-old Black American man reported, “I felt as if I did something wrong, despite me knowing that I didn’t do anything. I felt bad about myself” (p. 820). In addition to the emotional responses often referred to as fear and nervousness in the presence of police officers, helplessness is also a common outcome, as expressed by an 18-year-old Latino man also from the Nadal et al. study: “It happens a lot, so it really doesn’t matter” (p. 821).

The SQF approach was mentioned by two studies (Motley Jr. & Joe, 2018; Sewell, 2017) from different countries (the United States and England). These studies used police departments’ data on stops to assess whether this practice disproportionally targets Blacks. Common results showed a racial discrepancy in the frequency of stops. As officers register the citizen’s race according to their own prejudgments, however, it is important to note that Barnum and Miller’s (2014) results suggest such data can be influenced by factors such as the stop site being in a lower income neighborhood, where officers tend to report the person as Black.

We found no research reported in Portuguese or conducted with Brazilians or other Portuguese-speaking citizens. This may be due to the keywords being searched in only one database. Thus, the selection of a single database is a limitation of our review. Despite that, data regarding institutional racism in the Brazilian police can also be found. According to Cerqueira et al. (2017), the number of registered deaths occurring in police interventions in Brazil during 2015 was extremely high (3,320), even higher than the number of deaths after a robbery (2,314). Critically, during the same year, 71% of murdered people in the country were Black. There are no data regarding the percentage of Black people killed during police interventions, but one can expect numbers at least equal to that for total murders. The negative perception of police officers by Black people is also present in Brazil. For example, a survey showed that 68.3% of Brazilian respondents believed that the police acted in a racially biased manner (Instituto Brasileiro de Geografia e Estatística, 2011).

Future research should investigate traffic stops—which have proved to be a favorable environment for racial discrepancies in police procedures in the United States (Barnum & Miller, 2014)—in different countries, assessing the generality of this unfortunate phenomenon and its cultural particularities. Due to social and racial inequalities in Brazil, for instance, White people tend to have more financial resources, which influences the purchase of goods such as vehicles. In addition, cars are especially expensive in Brazil. Because we could not find studies relating racial prejudice and traffic stops in Brazil, researchers could investigate, for example, whether it is more common for police officers to stop Black people driving cars that are considered expensive, compared to White individuals, and whether Black drivers are stopped more than Whites.

The absence of studies in Brazil and other countries might also be due to the difficulty of accessing data on police stops. These data may not have been produced or made publicly available. Therefore, movements that demand constant production and publicity of data are paramount for researchers to be able to collect data that could help delineate public policies.

Behavior analysts have contributed to studies on racial prejudice and institutional racism (de Carvalho & de Rose, 2014; Mizael et al., 2016; Mizael & de Rose, 2017; Mizael & Sampaio, 2019). In that line, future research could investigate, for instance, the cumulative effect of prejudice in organizations such as the police force and some of the individual and cultural consequences that permit the maintenance of these behaviors, thus describing the components of the macrocontingencies involved (cf. Glenn et al., 2016; Mizael & Sampaio, 2019). These can help design cultural interventions aimed at reducing this and other types of prejudices—one thing that was unfortunately lacking in the literature we reviewed. Interventions to fight racial prejudice and institutional racism are badly needed, and a behavior-analytic perspective might just be what is missing for us to win this battle.

A possible way that behavior analysts could ameliorate racial bias in police stops could be by analyzing strategies already used, such as training programs. Generally, those types of programs focus on presenting the audience (police officers) with written and visual material describing how to better interact with civilians. According to Machado and Lugo (in press), this type of strategy could be effective to change verbal behavior. However, to establish specific skills (e.g., to use force only when no other alternative is possible), it could be more effective to use performance- and competency-based strategies. In this sense, behavior analysts could evaluate the use of video modeling or behavioral skills training to decrease racial bias in police stops.

Body-worn cameras (BWCs)—cameras attached to a police officer’s chest area or head that record their interactions with civilians—are another strategy already in use (see Voigt et al., 2017). According to Machado and Lugo (in press), they have been used by the U.S. Department of Justice since 2005, and, each year, more cameras are available to police officers. According to the authors, BWCs could be seen as an antecedent manipulation that could evoke socially congruent behavior (covert rules or overt behavior) or increase the probability of behavior that could avoid the punitive consequences related to racial misconduct (Machado & Lugo, in press). Those videos could also be used in sessions where a trainer reinforces professional behavior (e.g., by adding points that could lead to a higher salary) and where police officers are held accountable for unprofessional behavior, such as the abusive use of force (e.g., the police officer loses points that decrease their salary).

Racial bias in police stops is an urgent and fundamental matter affecting different aspects of society (e.g., health, traffic, imprisonment). Our review confirms the unfortunate and systematic presence of this social issue in different countries. We hope that such information can be useful in furthering more basic and applied research, helping to formulate interventions to reduce individual biases and to implement measures to decrease institutional racism.