INTRODUCTION
Talk of systemic racial bias in policing has been more prominent in the public forum in the past couple years but this doesn’t mean there has been a lack of research into issues of race and policing. Far from it, research in this area has been going on for decades, it’s now only recently that claims of systemic bias have become more vocal. But while there is ample evidence of racial disparities in police interaction outcomes, as we were taught early in graduate school, “disparities don’t equal discrimination”. So the question becomes, is there evidence of systemic racial bias in policing?
In this literature review, I can only just touch on the massive amount of literature examining racial outcomes in policing that has gone on in even just the last 25 years but I hope to provide a general overview of the issues of disparities, the arguments presented for systemic bias, the explanation of these disparities through the scientific research on arrests and use of force, and the problems with current methodologies that try to prove systemic bias.
RACIAL DISPARITIES DO EXIST BUT IS IT BECAUSE OF RACIAL BIAS?
The prevailing argument for systemic bias is the data showing clear disparities between Whites and Blacks/Hispanics in police interaction outcomes in areas like stops, use of force, and arrests. However, in the public forum this is played out with extensive coverage of Blacks who were killed in confrontations with the police, which then is portrayed as racial injustice and proof of systemic bias to the public. As Reinka and Leach (2017) noted in sociologist Weitzer’s 2015 contention that “the seemingly serialized nature of publicized killing after killing of an unarmed Black person every few months since Trayvon Martin’s killing in 2012 can easily be viewed as a pattern of systemic bias” These minorities are stopped, frisked, arrested, and have force used against them disproportionately in comparison to their population demographic and in comparison to Whites (Petersilia, 1983; D’Alessio and Stolzenberg, 2003; Ridgeway, 2007; Gelman et al., 2007; Tapia, 2011; Buehler, 2017; Balko, 2020; Tobias and Joseph, 2020).
For example Ramchand et al., (2006) reference a study showing Blacks are 2.5 times more likely to arrested for marijuana possession than Whites. Gelman et al. (2007) noted in their examination of NYC PD stop and frisk records that although Blacks and Hispanics are stopped more frequently than Whites, since only 1 out of every 8 White person who was stopped was subsequently arrested, compared to 1 out of 9 Hispanics and 1 out of 10 Blacks, that this demonstrates the police are stopping minorities indiscriminately and with less cause, than Whites. Cooley et al. (2019) even demonstrated that the racial disparities between Blacks and Whites in frisks, searches, arrests, and use of force had slight increases, between .3 and 1.7%, for people stopped in groups versus those alone, even after accounting for the discovery of illegal contraband. In their meta-analysis of probability of arrest, based on 27 independent data sets, Kochel et al. (2011) calculated that the odds of a Black person being arrested was 26% compared to a 20% chance of a White person being arrested. Lytle’s 2014 meta-analysis of suspect characteristics on arrest also found slightly higher arrest probability with minorities compared to Whites, with Blacks 1.4 times, and Hispanics, 1.25 times more likely to be arrested than Whites and non-Hispanics. Mitchell and Caudy (2015) examined racial disparities and drug arrest rates and after finding that the disparities could not be explained by differences in drug offending, non-drug offending, or residing in neighborhoods with heavy police emphasis on drug offending, concluded that racial bias is at work. Ross (2015) in an examination of racial bias in police shootings at the county level found that there was a 3.5 times higher probability of an unarmed Black person being shot by police than an unarmed White person being shot by the police. In a study of 2012 NYPD SQF data Morrow et al. (2017) found Black and Hispanic citizens were significantly more likely to experience non-weapon force than Whites
These disparities are seen of evidence as racist policing. Researchers like Richardson, (2010) and Pratt-Harris et al. (2016) claims that connection between race and crime is illusory and is produced by racist police officers, racist police practices, and other racially biased social institutions. Much of the promotion for systemic racism in policing comes from proponents of critical race theory (CRT) like Tolliver et al. (2016)and Tobias and Joseph (2020) by hearkening back to the country’s founding to make claims of a White supremacist basis of government and policing that was meant to subjugate Blacks and continues to this day. CRT proponents suggest that because of the difficult historical relationship between minorities and the police that policing is racist and populated by officers with racist attitudes,explaining that the police are simply a force to be used by rich White people to control an unruly Black populations (Roberts and Race, 1999; Pratt-Harris et al., 2016; Tobias and Josephs, 2020) and that the socialization that goes on in police departments have promoted the stereotype of Blacks as being violent criminals(Tolliver et al, 2016) and that police focus on minorities is in fact racist (Balko, 2020). Welsh et al. (2020), in their San Francisco PD traffic stop research, state that in interviews with officers that “officers’ accounts excuse, justify, or otherwise negate the role of race in routine police work” but yet despite a lack of explicit racism they see officers verbalize micro racial aggressions by adhering to racial narratives, which they say will then eventually lead to disparate outcomes. Rose (2018) claims the biased nature of violent police interactions are not held accountable in court because she presumes that racism extends to prosecutors reluctant to judge officers’ actions as wrong, to grand juries failing to indict, to juries failing to convict, and to judges failing to punish. Kahn et al., in 2016, found that based on use of force case files and subject booking photos, that police used less force with people who looked more “stereotypically White”.
OVERVIEW OF RESEARCH ON ARREST AND USE OF FORCE
However, simply because disparities exist in descriptive statistics or even in multivariate regression models is not evidence that there is racial bias or discrimination.. Descriptive statistics provide only raw numbers and exaggerate racial disparities and further analysis is needed to move beyond correlation and into causation (Ridgeway, 2007). Modeling with appropriate variables is important and can provides answers that less complete or nuanced models cannot. Multivariate statistical modeling is necessary to control for the influence of other variables that might drive the racial differences, such as variables like education, income, poverty, and social status, whose lower levels are both associated with Blacks and with greater likelihood of criminal involvement. Other variables like residency in high crime neighborhoods, crime rates, socially disorganized neighborhoods, and neighborhood gang activity help drive the amount of police presence and activity and these areas are also disproportionately populated by minorities. Because of the correlation, or association, that is shared between these variables and minorities like Blacks and Hispanics, they must be controlled for, in a sense, taken out of the equation. Without that accounting, it may give the impression that the disparities in outcomes like arrest are because the subject is facing racial bias, when instead it is because of the other criminogenic variables that some minorities are disproportionately associated with. Two areas of policing that have been claimed to be racially driven are arrests and the use of force. In the following sections, an overview of some of the scientific research on race in relation to these police activities is examined that runs counter to the argument of racial bias in policing because of refinements in defining statistical modeling and the inclusion appropriate variables
Arrests
While some commentators focus on racial disparities in arrest as evidence of racial bias, “social scientists remain unsure as to whether Black or White criminal suspects have a higher probability of arrest” while other research finds no significant association between a suspect’s race and the likelihood of arrest (Arndt et al., 2020). Often there are other variables that once accounted for allow those differences in probability to shrink or disappear and those racial disparities can be explained. While evidence of disparities have existed for decades so has research that examines the reason those racial disparities appear to exist. Disrupted or disorganized social structure, poverty and lower social status can confound racial data on arrests as minorities disproportionately reside in social disorganized and poor neighborhoods. For example both Hollinger (1984) in examining a DUI crackdown and, Sealock and Simpson (1998) in their research on Philadelphia juvenile offenders found the racial effects on arrest lost statistical significance once socioeconomic status (SES) was accounted, indicating a person’s social status, rather than their race, became the significant factor in whether a person was arrested. Fielding-Miller et al. (2020) in studying whether high proportions of drug arrests for Black males were linked to areas of higher levels of White female population found that the connection was attenuated, though not completely eliminated, by including socioeconomic variables in their model concluding that economics must be included in the discussion of criminal justice disparities, not just race alone. In 2008, Kirk also found that Chicago minority youths faced significantly higher levels of concentrated poverty and lower levels of collective efficacy than White youths and that unstable family structure explains much of the disparities in arrest across race and ethnicity. They conclude that while substantial arrest differences still existed, numerous family level covariates are significantly and substantially associated with arrest and the racial and ethnic differences in family demographics and structure explain large percentages of the disparities in arrest between Black youth and other racial and ethnic youths. Finkeldey and Demuth (2019)looking for a connection between self-identified race/ethnicity, self-identified skin color, and self-reported arrest found that focusing on race or ethnicity masks difference in arrest by color but also found that social disadvantage partially explained those associations. In a youth survey, Schleiden et al. (2020) found partial support for a social disorganization hypothesis, determining that Blacks experienced more neighborhood disorder than Whites, which was linked to their young adult arrest rates and it was Blacks exposure to violence that was linked to more alcohol use, delinquency, and arrests in emerging adulthood.
Other behaviors specific to the conditions minorities may live in may be exhibited that expose minorities to a greater likelihood of arrest. Camplain et al. (2020) in their study on drug and alcohol arrests discovered that minorities were more likely than Whites to be booked into jail on both misdemeanor and felony charges alcohol /drug charges than Whites and it cannot be attributed to their greater drug or alcohol use in general. Mitchell and Caudy (2015) arrived at a similar conclusion suggesting disproportionate minority drug arrests are not explained by drug offending, non-drug offending, or residence in neighborhoods with a high police focus on drugs, but instead attribute it to racial bias. However, back in 2006, Ramchand et al. revealed that one likely explanation for part of the disparities is not use but the difference in purchasing behavior between Blacks and Whites. Their analysis showed that Blacks, as compared to Whites, were nearly twice as likely to buy outdoors, three times more likely to buy from a stranger, as well as significantly more likely to be buy away from their homes. These factors afford more police scrutiny and an increased likelihood of arrest.
A number of other factors, once included in statistical models, provide evidence that racial disparities can be explained by other variables, rather than support the suggestion that police officer behaviors are driven by racial bias. One of the most basic variables that needs to be accounted for is differential criminal involvement of different racial groups. Petersilia (1983) in her examination of arrest in California from 1980 found mixed findings with Whites being somewhat more likely than minorities to be arrested on a warrant and considerably less likely to be released with charges. While noting that the actual of likelihood for any offender to arrested is quite low, regardless of race, she found that Blacks were not overrepresented in their race relative to the kind or amount of crimes they commit. As Kirk (2008), Tapia, (2011),Schleiden et al. (2020) and other researchers have noted through official data, the National Crime Victim Survey, and self-reports that Black and Hispanic youths disproportionately commit serious crimes relative to Whites. D’Alessio and Stolzenberg (2003), utilizing National Incident Based Reporting System (NIBRS) data, the successor to the UCR, examined 335,619 incidents of rape, robbery, and assault in 17 states during 1999 and determined the likelihood of arrest for White and Blacks suspects was roughly equal and that the “disproportionately high arrest rate for Black citizens is most likely attributable to differential involvement in reported crime rather than racially biased law enforcement practices”. Schleiden et al. (2020) while not finding a difference in frequency of delinquent behaviors between Black and White youths, found significant differences in self-reported offenses with Black youth disproportionately being involved in buying, selling, or holding stolen property, deliberately writing a bad check, breaking into a house to steal something, using a weapon to get what they want, and using a weapon in fight. Despite these findings, Schleiden et al. characterize the higher arrest rates for Black youths as possible evidence of racial bias. They did not investigate whether the neighborhood disadvantage and exposure to violence for Blacks could lead to differentiated criminal behavior, for example more serious, or violent behavior, that would more likely result in arrest, as opposed to examining just frequency of behavior.
In that regard, another variable that has to be accounted for is seriousness of the offense, whereas the more serious the offense, the greater the likelihood of arrest.Schleiden and her colleagues (2020) had already noted the research that indicates minorities are committing more crime over a longer period of time, are engaging in more criminal behaviors, and committing more serious crimes than Whites, which consequently are more likely to be reported to the police , and thus more likely to result in arrest. Sealock and Simpson (1998) as well found that officers strongly considered offense seriousness in their decision to arrest. Brown (2005) also found that regardless of suspect race, legal factors like offense serious influence the decision to arrest. The seriousness of offending was also a factor in Tapia’s 2011 national youth survey. Noting research that indicates Blacks’ self-reported serious offense rates are double that of Whites, Tapia found a disproportionate amount of minorities in gang membership and that delinquency and arrests rates are much higher for gang members than non-gang members. However their regression analysis showed that as gang members they had increased rates of arrest, but their race was not a significant factor in arrest. Tapia concluded that because minority youths make up a much larger portion of gang membership they feel the effect of arrest greater. This likelihood of arrest is affected by the introduction of criminal behavior, not by the race of the gang members themselves. Raphael and Rozo (2019) also found that offense severity and criminal history were significant factors in arrest. While Black juvenile offenders were booked at a higher rate, a significant portion of the Black/White disparity was attributable to differences in arrest offense severity and criminal history.
Another legal factor that also strongly influences the likelihood of arrest is the presence of evidence or solvability factors.Brown (2005) found that regardless of race, the presence of evidence was a significant factor in the decision to arrest. Briggs and Opsal (2012) in examining victim ethnicity on arrest in violent crimes found that while race and ethnicity of the victim are related to police clearance of cases, the relationship between case clearance and case solvability factors is stronger. Arndt et al. (2020) also found that in homicides, suspect race did not play a noteworthy role in influencing the likelihood of arrest once the strength of physical evidence against the subject is accounted for. Solvability and the presence of evidence, as well as arrest itself, may be linked to demeanor of the individuals involved either as witnesses or suspects. Extensive research has shown that minorities tend to hold more negative views of the police. These negative views can prompt uncooperative behavior from victims and witnesses or behavior that is confrontational or resistive, which will influence the likelihood of arrest. However, Drawve et al. (2014) in their analysis of the likelihood of arrest for aggravated assault, found that Blacks had a significantly lower chance of being arrested than Whites, with the authors postulating that the decreased likelihood of arrest for Blacks may stem from Black witnesses and victims lack of cooperation with the police, or perhaps by less of a focus by police on Black crime linked to this lack of cooperation. Other researchers like Brown (2005) have also found that suspect demeanor plays a role in the likelihood of arrest. More than a negative demeanor may develop from vicarious or actual interactions with the police. Gibbons et al. (2020) in their longitudinal study of Black young adults found that those who perceived to have been racial discriminated against, including being “hassled” by the police, especially in their childhood, were more likely to have reported being arrested or incarcerated as adults. However, in the path to that connection were also self-reported deviant associates, substance use, and illegal behaviors that in all likelihood contributed to being “hassled by the police” and their subsequent higher arrest rates. They also found that those who exhibited high racial pride were less likely to commit illegal acts in general but yet reported more illegal behavior after perceived racial discrimination from the police.
Studies need to account for policing styles, practices, or policies that while not driven by racial bias, may disproportionately affect minorities. Drawve et al. (2014) notes that overall police strategies and resources will affect the likelihood of arrest. Police resources will undoubtedly be focused in areas of greater need as well as for particular efforts that focus on DUIs, street crimes, or drug activity. Crackdowns, problem oriented policing, and hotspot policing have all demonstrated effectiveness on reducing crime When these efforts are occurring in areas with large proportions of minorities, such as high crime areas, minorities will be affected by stops and arrests disproportionately (Weisburd et al, 2016) though Schleiden et al. (2020) fear that will affect officers’ perception of differences in behavior and character between Blacks and Whites. In examining the racial disparities found in the NYPD stop, question, and frisk practices,Evans et al. (2014) found the police focused those activities in areas with high levels of violent crime, which are areas that are highly populated by minorities. Lum (2011) describes that policing is moving away from a person based approach to one based on place. Various factors in areas or neighborhoods help drive criminal activity and disorder and addressing those underlying place based problems is at the core of community based policing, as well as utilized in hotspot policing and anti-gang and drug interventions. These places and their characteristics such as SES, poverty, racial and ethnic makeup, disorder, crime, pedestrian and traffic density and land use provide place-based cues which will affect officer behavior, decision, making and attitude (Lum, 2011) as well as affect precinct and department policy and behavior. She notes while they have proven to be effective as anti-crime efforts they’ve also drawn criticism because police presence in minority areas results in disparate arrest outcomes. In her place based study of the interaction of the proportion of race, level of wealth, and level of violence in an area, the more violent an area accounted for greater arrests but she also found the police more informally handled complaints (less likely to write reports or make an arrest and reducing the seriousness of crime classifications) in areas with greater wealth and in areas of larger Black populations who are more disadvantaged. Lum unfortunately wasn’t able to determine the reason for these opposing results, considering officer attitude or beliefs, department or precinct policing style, amount of police resources, community demands or expectations for service, or an interaction of all of them may have accounted for the results. Raphael and Rozo (2019) found that while there were racial disparities in non-violent felony youth arrest bookings, a very large share of that difference was accounted for by agency variables. Agencies that tended to arrest Black youth were equally likely to arrest Whites and Hispanics. Tougher policies regarding arrest in large cities will also affect minority arrests as these cities have higher minority populations. They note that 75% of the disparity in bookings between Blacks and Whites and 66% of the disparity between Hispanics and Whites were explainable by the observable factors of criminal history offense severity, and the penchant for certain law enforcement agencies to arrest and formerly book youth offenders
One area not frequently examined, typically for lack of data, is officer race in the decision to arrest. Brown and Frank (2006) note that some efforts in police reform have focused on increasing the number of Black officers with little empirical evidence that an officer’s race or ethnicity is actually related to their behavior toward citizens. In their systematic observation study in Cincinnati, they made an interesting finding. In general, White officers were more likely to arrest suspects than Black officers, but Black suspects were more likely to be arrested when the decision maker was a Black officer.
Unfortunately, not all researchers in preparing a study seek out, have access to, or utilize data from these explanatory variables. The types of data will significantly affect the quality of the findings. For example, Penner and Saperstein (2015) used the US National Longitudinal Survey of Adolescent Health to “disentangle the effects of self-identifying as Black compared to others identifying you as Black on subsequently being arrested”. They revealed that the odds of arrest are three times greater for someone identified as Black, proclaiming that racial perceptions play an important role in arrest disparities. However they also admitted that they had no detailed information on the circumstances surrounding the arrests such as evidence, demeanor, or compliance. Other variables that were not included would also have been informative (criminal history, police record, type of offense) as these might also have factored into the arrests. A study like this has little social science value and amounts to little more than an observation as it seeks to determine a component that might have affected arrest, without actually looking at the arrest occurrence itself. Other studies as well face these limitations. In the previously discussed 2008 Kirk study, he notes the lack of data on victim/offender relationship, seriousness of the offense, and subject demeanor would all be factors related to arrest decisions. In preparing his 2011 examination of youth gang arrests, Tapia having discovered insignificantly higher rates of arrest for minority gang members compared to White gang members, also recognized limitations in the study that didn’t account for the complainant’s demand for arrest, presence of evidence, and subject demeanor which all affect the decision to arrest.
Other sociological reasons that lead to higher arrest issues that should be accounted for in samples include substance use and mental health issues, (Schleiden et al., 2020) which will likely have a disproportionate affected on disordered or poor neighborhoods, which will also likely be high crime neighborhoods with large populations of minorities. They emphasize that it is crucial that in examining racial disparities “to simultaneously consider multiple theories and include a variety of individual and contextual factors within the analyses”. However, Lantz and Wenger (2019) also demonstrated that researchers must approach their overall analytical approach carefully. They stated that in their quasi experiment examining potential differences in arrest for Black and White co-offenders, using traditional logistic regression, Black offenders were 75% less likely than White offenders to be arrested. Putting the data into a multi-level model with partnership data, including victim and offense characteristics, nested with demographic and arrest data, in examining within partnership differences between Black and White co-offenders they found that Black offenders were 3% more likely to be arrested than their White co-offenders. While the authors suggest that traditional regression analysis may contain significant selection and omitted variable bias, they do not provide an explanation about why the disparity may exist. Elliot (1995) even notes using arrest statistics to show anything representative about the general population or police activity is flawed. While many researchers are aware that arrest data may include false positives and errors of omission, it should be noted that arrest statistics may not be representative of offenders in the population. Elliot states that compared to self-reports of violent crime, arrest data underreports these offenses and research results are mixed on whether there are any racial differences in the number of self-reported offenses. Ultimately Lum (2011) determines that the only way to understand why the disparities occur is to continue to use further systematic and qualitative approaches including social observation, ethnographic analyses and in-depth interviews, and psychological examination of both officer and citizen mentality.
Use of Force
The same issues surrounding research into racial disparities in arrest exist for those disparities in the use of force. Researchers like Tolliver, et al. (2016) contend that the U.S. has a White supremacy basis and a racist adherence to the stereotype of Black males as violent criminals. They contend that White privilege is the cause of higher incidence of use of force with Black subjects. For proof, they reference evidence taken from shoot/don’t shoot videogame studies that a shooter bias exists because some results indicate Blacks were shot faster than Whites. The disparities witnessed between Blacks and other races in police use of force are often attributed to racial bias and discrimination (Buehler, 2017; Hehman et al., 2018;Ross et al., 2020; Durlauf and Heckman, 2020) For example, Pratt-Harris et al. (2016)contend that the core of policing is racist, incorrectly attributing policing’s roots to American slave patrols, and that Blacks disproportionate involvement in crime is untrue, blaming crime in those neighborhoods on racist city policies. They believe a stronger focus should be put on revealing the racist aspects of policing and that focusing on individual “bad apple” officers dilutes the message of systemic racism.
However, as with arrests, alternative explanations rather than racism exist as to why these disparities occur. Fridell and Lim (2016) note that neighborhoods matter; with more force used in disadvantaged, high crime neighborhoods. While these places are typically populated with minorities, once controlling for neighborhood context, the disparity in racial use of force disappears. Lautenschlager and Omori (2017), found a different neighborhood effect. While they found that use of force was concentrated in Black neighborhoods, neighborhoods with higher ethnic and racial diversity have decreasing force incidents but with increasing severity. Ross (2015) found that racial bias was most likely to occur in large metropolitan counties with low median income and a sizable portion of Black residents, especially with high financial equity, however he did not find racial bias in shootings to be associated with county crime rates.
It is well known and easily understood that serious crimes and crimes involving weapons or violence will more likely involve the use of force then other incidents. Arrest and complaint data have consistently shown a significantly higher degree of involvement of Blacks in these types of offenses like robbery, aggravated assault, and murder, which then drives disparities in use of force data. Research has indicated that use of force is related to racial/ethnic minority involvement in criminal activity and their resistance to police intervention (Fridell and Lim, 2016). Cesario et al. (2019) examined two year worth of officer involved fatal shootings. When adjusting for crime rate and minority criminal involvement they found no evidence of anti-Black disparities in fatal shootings, fatal shootings of unarmed citizens, and fatal shootings involving misidentification of harmless objects.
Subject demeanor compliance, and resistance strongly influence the application of force (Engel et al., 2000; Fridell and Lim, 2016; Mears, et al., 2017; Morrow et al., 2017) examined use of force in Terry stops and while Black and Hispanic had small but significant effects on the use of non-weapon force, many other factors had a much greater influence on whether force was used, including being frisked, arrested, matching subject description, engaging in violence, and noncompliance. In weapon use of force, race was no longer a significant factor but rather noncompliance was strongly related to that use of force, along with matching subject description, suspected of a violent crime, presence of weapons, being searched, frisked, proximity to offense and whether it was radio dispatch. High crime areas though had significantly lower use of weapons use and the racial makeup of the police precincts had no significant influence on whether force was used. Fryer (2019) found that in non-lethal force, while Blacks and Hispanics were 50% more likely than Whites to experience force in a police interaction, once situational context and subject behavior was accounted for no racial differences were found. He also found in officer involved shootings there were no racial disparities present. Atiba Goff and Barsamian Kahn (2012) suggest that officers have two kinds of authority, moral and physical. When subjects portray officers as illegitimate because of a presumption of racism, officers are stripped of their moral authority in the eyes of the subject, and thus themselves, and have only physical authority to rely on for compliance. This reliance on physical authority, generated from the situational context, may then generate racial disparities in the use of force, rather than actual racial discrimination.
Linking Implicit Bias to Racism in the Use of Force
Drakulich, et al. (2020) state that modern racism rests on two key principles: a denial or minimization of contemporary racial discrimination and inequalities, and a focus on individualism and meritocracy that (implicitly or explicitly) blames racial disparities on people of color by arguing that they lack the proper work ethic and discipline to succeed in today’s fair, equal-opportunity society. Levinson and Young (2009) declare that racism has been shifting away from overt and intentional racism of the past to covert and unintentional. They complained that in attempting to proving a case of racism; showing intent, and not just an adverse effect, is difficult and we should focus more on whether anything symbolic of racism was present rather than an actual intent to discriminate. They contend that racism can be non-conscious and automatic and that these biases can be uncovered by using the Implicit Association Test (IAT). Holroyd (2015) claims that because implicit bias exists, then if must be causing police officers to behave in a racially biased manner, and if it affects police officers then it must also affect policy makers in the department causing them to be racially biased as well, thus supporting his contention of institutionalized racism. Holroyd both calls for studies to examine if and how implicit bias affects policing, and yet proclaims that implicit bias is causing discriminatory police work. He states that raising awareness of implicit bias can change it, and that the effects of implicit bias must be mitigated because evidence of implicit bias mean policing is racially biased and thus illegitimate.
The Implicit Association Test basically measures the speed of a pushbutton response in an image/word association test. An image type, for example those indicating different races or genders, is paired with words that either have a positive connotation, like happy, or a negative connotation, like disgust, and participants are measured on their speed in correctly responding. The image and word types are then reversed and participants are measured again. If an individual associates a negative word faster with one image type than another it suggests they hold a bias or adherence to a stereotype. Humans tend to categorize people and object in to group membership, attributing stereotypes to those members in order to make sense and add structure their world. These implicit biases operate outside of conscious awareness and are not necessarily based in animosity as research has shown that implicit biases exist in people who consciously hold no prejudiced attitudes (Fridell and Lim, 2016). Fridell and Lim also note that people can override their implicit biases and have controlled responses. By recognizing their biases, motivated individuals can implement bias-free behavior (2016).
Tests like these demonstrated that people have an inherent bias toward their ingroup (the type of people that resemble them the most) and bias against outgroups (individuals who do not resemble them) and that these biases occur within all people in areas of gender, race, sexual orientation, religion, and body shape (Fridell, 2016; Fridell and Lim, 2016). In that regard, people naturally will be quicker to make positive associations toward people that look more like them and slower to make those positive associations toward people who appear different to them. Likewise, people will naturally be slower in making negative associations about their ingroup and faster in making negative associations about outgroups. Fridell and Lim also note however, these tests can also demonstrate the possible degree that people adhere to negative stereotypes about race and possibly indicate a racial bias in an individual they were perhaps unaware of or preferred to outwardly conceal (2016).
Commentators and researchers state that a persistent stereotype of the violent criminal as a Black male may generate bias against Blacks and may prompt the use of force, or greater use of force than would be necessary, as compared to police reactions to other races. Wilson et al. (2017) found in college student studies that photo examples of Blacks were subjectively assessed to a very small, but significant, degree, as having greater size, strength, or formidability of threat. However, in 2019, Johnson and Wilson found that while race does impact judgments of size and strength to a certain degree, raters primarily tracked objective physical features. They even found that in some cases racial stereotypes actually improved group level accuracy as these stereotypes aligned with nationally representative data of racial group differences in size and strength Kahn and her associates, in 2016, examined subject’s booking photos and coded them for the presence of typical Black or White facial features while examining the severity of force used in the arrest. In their regression analysis the police used less force with highly stereotypically featured Whites and they suggest an intragroup bias from White officers is a protective factor in the severity of use of force.
However while there are stereotypes of Blacks as dangerous, violent, or criminals, other stereotypes that citizens have about policing exist as well. Mears et al. (2017)states that both police and citizens can bring their biases to police encounters and that when minorities express negative views of the police, question police legitimacy, and believe their neighborhoods are subject to coercive policing, Blacks and other minorities may still perceive professional and courteous demeanor from the police as discriminatory (2017). Both direct and vicarious experiences influence minorities’ views of the police and their behavior, and negative perceptions of the police by minorities have been longstanding (Tapia, 2011). This will lead to noncompliance and resistance, and prompts the use of force from officers. Mears (2017) also notes that in encounters with subjects who are defiant, disrespectful, or violent may also mold officers perceptions and biases as well. Cooley et al. (2019) demonstrated in two studies that after merely reading about White privilege, study participants had increased perceptions of racism in violent encounters between the police and Black men, indicating the perception of racism in police actions is subjective and subject to outside and vicarious influences. Levin and Thomas (2005) in a study of how racial identity affects perceptions of police brutality, produced videotapes of a police interaction, changing the race of the officers in the different recordings, and found that both Black and White participants were significantly more likely to identify identical behavior in police officers as racist when both officers were White. Hall et al. (2016) suggest, without any supporting evidence, that police officers are high in Social Dominance Orientation and are thus more likely to inflict severe punishment on those who don’t submit to there demands. They also claimed, again without evidence, that police officers typically have high conformity and low universalism values and, similarly to the non-officers participants in their study, would rate failure to comply more seriously, are more supportive of harsh punishment, and have less sympathy for offenders. The authors suggest that it’s the officers’ desire for dominance that prompts them to use “resisting arrest” or “failure to comply” as a precursor and justification to employ lethal force when their demands aren’t met.
Research to determine if implicit bias affects behavior typically took the form of video shooter simulations. Participants watch a video portrayal and must correctly identify whether the Black or White target in the video is armed with a weapon or holding a harmless object and push either a “a shoot” or don’t shoot” button. Participants are timed in the speed of their responses and their accuracy in decision making. Results of these types of studies revealed some consistent results (Correll et al. 2007). Participants of all races were quicker to decide to shoot armed Black targets than armed White targets and decide not to shoot unarmed White targets faster than unarmed Black targets (Correll et al., 2002; Correll et al., 2007; Correll and Keesee, 2009; Sadler et al., 2012; Senholzi et al., 2015). This suggests an adherence to the stereotype that Black males are violent or dangerous and the magnitude of the bias varied based on the perceptions of the cultural stereotype but not with personal racial prejudice (Correll et, 2002). However, when comparing police officers to citizen participants, officers also were quicker to decide to shoot Black subjects, but while citizens tended to favor a shoot response for Black subjects, suggesting bias, officers, who’s racial bias in response times did not differ between White and non-White officers, were significantly more accurate than citizens in making the right choice and showed weak or non-existent racial bias in their shoot/don’t shoot decisions (Correll, 2007; Correll and Keesee, 2009;Sadler et al., 2012, Atiba Goff and Barsamian Kahn, 2012). James et al. (2014)wanted to explore the possibility that the racial biases exhibited truly affected behavior and questioned whether those type of video shooter studies would accurately predict real world behavior as the tests differ significantly from a real life situation. The researches utilized general population participants in a highly realistic deadly force shooting scenario and found, similar to other research, that associations between race and violence did not translate to biases in shooting behavior.
Implicit bias is not universal or immutable. Individuals can control their implicit biases through implicit bias awareness, cognitive control, and training, (Richardson, 2010) which may be available to some officers as well as conventional officer training, that while not affecting response time on stereotype-incongruent subjects (unarmed Blacks and armed Whites), did improve accurate responses over those of a citizen (Correll, 2007; Correll and Keesee, 2009;Sadler, 2012; Atiba Goff and Barsamian Kahn, 2012; Correll, et al., 2014). Fridell and Lim (2016) also note that reaction time biases amongst officers were not universal, with bias large for officers from large cities, those with high minority or Black populations, and for officers who perceive greater violence in their communities. As members of a community, officers may share the same biases of their community or area and it’s critical to examine these biases in the communities they serve and how these might influence officer attitudes and the implementation of police policies(Correll, 2007).
Plant et al. (2011) also explored the little researched interaction of gender and race influencing the use of force. Using a shoot/don’t shoot video simulator in a study examining gender, White participants were biased against shooting a White female compared to a White male. In a second study, White participants showed a bias toward shooting Black men but a bias against shooting Black women and White ingroup members. The authors indicate that not just race, but gender, and its connection to perceptions of violence, is an important factor in shootings that have a racial component. Sadler et al. (2012) also explored biases toward other groups than Blacks. Interestingly, while college students participants only showed a bias against Blacks in their reaction times, officer participants also demonstrated reaction time biases toward Latinos relative to Asians and Whites, and toward Whites relative to Asians; racial biases in their reaction times that mirrored perceived racial criminal involvement. They did find that the more aggressive police officers perceived the stereotype of Blacks to be, the more accurate they were in decisions about whether Blacks were armed. While increased perceptions of aggressiveness in Asians led officer to be more accurate in their decisions, they were less accurate with Latinos when having holding perceptions of aggressiveness. Participants were also better able to distinguish between weapons and non-threatening objects better when held by Black and Latino subjects compared to Whites and Asians. The authors consider that the association of Blacks and Latinos with danger may lead to faster correct responses through increased cognitive attention to a potential threat but these biases in reaction times did not manifest themselves in the decisions themselves and no racial bias was evident in officers’ decision-making.
The influence of a potential threat doesn’t have to be based in race, but in the perception of the level of threat based on contextual cues or how attentive one is to potential interpersonal threat.. Correll et al. (2011)argue that the bias seen in faster shoot responses to armed Black targets and faster don’t shoot responses to unarmed White targets may not stem from racial animosity but reflects the perception of threat, specifically the threat associated with Black males. To examine the influence of danger cues other than race, the researches used a shoot/don’t shoot video simulator with the targets embedded in threatening and safe backgrounds. While a typical racial bias was seen in decision making toward Black and White targets in safe backgrounds, threatening backgrounds saw this racial bias disappear as the dangerous contextual cues provided more influence on shooting behavior, with the reduction in racial bias largely due to an increased tendency to shoot White targets. . However, Miller et al. (2012) suggest that the basis of shooter bias may not involve racial or cultural stereotypes. In one study, using arbitrarily formed groups not based on racial or cultural stereotypes, and in another, using Asians as a group typically not stereotyped as dangerous, they found that participants with strong beliefs about interpersonal threats were more likely to mistakenly shoot outgroup members than ingroup members.
Different neural mechanisms underlie this increased cognitive attention to threat assessment based on the stereotype of the violent Black male. Senholzi et al. (2015) conducted a shoot/don’t shoot video simulator study with participants connected to an MRI. They exhibited the typical bias toward quicker shoot response for Blacks and quicker to not shoot for Whites but different parts of the brain were activated depending on the race of the target. Greater implicit Black-danger associations were associated with increased activity in the amygdala (the portion of the brain responsible for the fight or flight response). The parietal and occipital regions (responsible for processing movement and vision information) also exhibited greater activity for armed Blacks than Whites and a greater connectivity of these regions to the amygdala with armed Blacks. However, the anterior cingulate cortex (an area of the brain that detects and monitors for error and suggests an appropriate motor reaction) was preferentially engaged for unarmed White targets over unarmed Black targets.
Correll et al. (2014) notes that while target race does affect officers’ reaction times they generally do not show a pattern of biased shooting, suggesting that police performance depends on the exercise of cognitive control which allows officers to overcome the influence of stereotypes. Mears, et al. (2017) also expands on the cognitive processes that goes into these decisions. Police/citizen encounters require rapid assessment that demands reliance on cognitive shortcuts, “thinking fast”. When rapid mental processing is needed, as in a dangerous situation, all people rely on cognitive short cuts to make the quickest, most accurate assessments in a situation. However if those shortcuts are based on a flawed assumption they can lead to decision making errors. Ma et al. (2013) consider though that officers must have ample cognitive resources to fully regulate automatic response, and thus control implicit bias. If this cognitive control is compromised through fatigue it may lead to great racial disparities in shoot/don’t shoot decisions.Similarly, James (2018)in examining the stability of implicit bias in police officers found that implicit bias was not stable and that when officers slept less prior to testing they demonstrated stronger associations between Blacks and weapons. Nieuwenhuys et al. (2012)also explored the effect of anxiety on officers’ decision to shoot. By utilizing low and high anxiety physical experiment conditions, the researchers found that officers in anxious conditions made more mistakes in correctly shooting only armed targets, had reduced shot accuracy, but also responded more quickly when the target had a gun. With gaze behavior not changing, the authors concluded in anxious conditions officers were more inclined to respond on the basis of threat related inferences and expectations than relying on visual information.
Research Issues in the Use of Force
Methodological issues and disagreement between researchers complicate the question of racial bias in the use of force. While Ross (2015) found that Blacks had a slightly lower probability of being shot if armed vs unarmed compared to Whites and Hispanics, unarmed Blacks had a 3.5 greater chance of being shot than unarmed Whites. However, Ross’ study did not include important variables like level of compliance or resistance, type of weapons subjects were armed with, whether threats were made, the seriousness of the offenses, and encounter rates. Despite the authors stating that they can’t know all the circumstances involved or the psychology of the officers at the time, they suggest the disparities exist because of racist norms in police departments. However they ignore clear evidence supporting community violence as a factor as well as evidence of Social Disorder theory. Other study design and analysis problems exit as well. Legewie (2016) examined the rate of change in use of force incidents following the line of duty deaths of NY police officers. However, while the pretest period gathered data for a year, the post test period following the officers deaths was only two weeks. Legewie states that there was a significant increase in use of force against Blacks following two Black incidents, there was not a significant increase after a White and Hispanic incident, and no evidence of any institutional bias. However the small sample and a lack of examination of contextual issues lessens the quality of the study. Such things as, three of the four suspects were killed or arrested shortly after the incidents, the attacker still on the loose was Black, one of the Black attacks was an ambush, and that Blacks contributed to more than one death may have influenced officer actions based more on the circumstances and less on race.
Fridell (2016) notes that seven studies on race and police use of force came out in the first half of 2016 that were announced in the media that paraphrased them ranging from “there is bias in the use of force”, to :there is bias in some types of force, but not others”, to “there’s no bias in the use of force”. She explains that these disparate findings hinge on the methodologies used; sample characteristics, the number and types of agencies examined, how concepts were operationalized and the number and types of variables used. The quality of the study making a claim is important and research consumers need to critically analyze the results of studies. Rivera and Ward (2017)also make the case for improving analytic frameworks and utilizing an integrated approach in the study of race and policing in both methodology and evaluative analysis. Perspective, approach, and study quality all matter. Differing methodologies and analyses, with varying degrees of rigor, can produce disparate results, with no clear research consensus. Hollis and Jennings (2018) conducted a meta review of 41 studies which examined public and police officer perceptions of force, rates of use of force, types of force used, neighborhood context correlates of use of force and the severity of the force used. In the end they concluded that the relationship between race and use of force “remains unclear”‘ This was the conclusion Petersilia reached in 1983, when 30 years of prior research still did not have a definitive answer. Fryer (2018) states that at while there are racial differences in the use of non-lethal force, in regards to potentially lethal use of force the most granular data shows there is no bias in police shootings. However Durlauf and Heckman (2020) take issue with that conclusion, contending that just because statistically no difference was shown between races in police shooting, doesn’t mean there wasn’t racial discrimination. Ross et al. (2020) also argue against the conclusions of Cesario et al. (2019) who stated that with proper benchmarking, data analysis shows no racial bias in shootings. Ross et al. argue that Cesario et al.’s benchmarking of crime rates doesn’t properly control for crime rate differences but instead masks true racial disparities in the killing of unarmed people by the police. Ross argues that unarmed individuals who were killed should be benchmarked as a ratio of the “noncriminal population”, simply assuming that not being armed, meant you were not aggressive, and thus not a criminal. Taking those assumption, Ross states it would show there were racial biases. Roussell et al. (2017) takes issue with a James et al. 2016 study that demonstrated that officers had slower response times and fewer shooting errors in a video simulation. James and associates suggested that a fear of adverse legal and social consequences led police officers to be more cautious in shooting decisions involving Blacks than Whites, dubbing this a “reverse racism” effect. Roussell et al. take offense at this characterization, stating reverse racism can’t exist, the term is used to undermine efforts toward racial equity, that if officers did fear repercussions they must realize how unlikely it would be that they are punished, and criticized James et al. for not properly considering that studying police and race is actually about studying racism and policing.
MAKING SENSE OF DISPARITIES
Atiba Goff and Barsamian Kahn, in 2012, examined why we know less than we should in regard to racial bias in policing. While research has found evidence of disparities, there is a lack of evidence establishing that racial discrimination exists in the social institution of policing. Public opinion data about the presence of systemic bias, often based on disparities, carries no value as it offers only subjective views and no evidence. Distinguishing between disparity and discrimination can be complicated by a number of factors. Limitations in department data, subjective claims of racism without witnesses, limitations in national data on crime and race, and other factors make it difficult to distinguish between simple racial disparities in policing and racial discrimination at the officer, local, or national level. In 1983, Petersilia discussed the reasons for disparate results and the difficulty of distinguishing racial discrimination from racial disparities, reasons that still exist today.. Some studies utilize databases that are too small to permit generalization, others failed to control for enough (or any) of the other factors that might account for apparent racial discrimination. Atiba Goff and Barsamian Kahn (2012) discuss finding the appropriate benchmark for the variables used in analysis can be problematic. For example, in research on racial profiling or racial disparities in stops or arrestees, trying to find an accurate ratio, essentially getting the proper denominator, is crucial, and different benchmarks can provide disparate results. In their analysis of pedestrian stops, using the population of crime suspect descriptions, Blacks pedestrians were stopped 20 to 30 percent lower than their representations in crime suspect descriptions. But Blacks were stopped at nearly the same rate as their representation in the arrestee population. However in using the least reliable (as it doesn’t take into account differential rates of crime participation by race or for differential exposure to the police) but more widely used census data, Black pedestrians were stopped at rate 50% higher than their representation in the census.
Atiba Goff and Barsamian Kahn (2012) state while some research puts police officers racial bias at roughly the same level as the public in general it does not show that theses biases translate to discriminatory behavior, noting that in social psychology attitudes traditionally predicts less than 10 % of the variance in behavior. They note that even if racially biased attitudes produce some level of biased behavior it’s not clear how much it produces. Ridgeway discovered in 2007 that in the NYPD SQF program just 7% (2,756) of the total number of officers accounted for 54% of the total number of 2006 stops. In patrolling the same areas, at the same times, and with the same assignments a very small percentage of officers (15 total) stopped substantially more Blacks or Hispanics than other officers. Internal benchmarks are necessary to compare officers performance to similar officers to see if any discriminatory patterns exist but these can measure only officer behavior, not institutional behavior (Ridgeway, 2007; Ridgeway and MacDonald, 2009; Atiba Goff and Barsamian Kahn, 2012) However this would provide a statistical means to identify the “bad apples” in the department, important as these bad apples are possibly performing racially biased police, affecting those they interact with, as well as continue to reinforce the arguable narrative of racist policing (Futterman et al., 2007). Goff & Kahn also recount the problems of a lack of access to police data through secrecy or because it was uncollected, and a difference in police and academic culture that can result in a poor collaboration as well as problems with methodology and rigor. Using low quality data and focusing on base rates and descriptive statistics, research like this fails to utilize multivariate analyses which can isolate the roles of the institution and officer to examine their influences. With these limitations in demographics, policies, and outcomes, it is difficult to develop principles and theoretical frameworks with which to predict racial disparities and diagnose racial discrimination.
CONCLUSION
While the current social narrative beats the drum of systemic racial bias in policing, research tells a different, more nuanced story. It is a well-researched and undeniable fact that racial disparities exist in police/citizen interactions. But regardless of the opinions of social commentators and some researchers, these disparities are in no way proof of racial discrimination. Raw data and descriptive statistics exaggerate the real racial differences in police activities. Researchers need to locate and utilize appropriate benchmarks. Discrimination can’t be determined if the ratios that would indicate that conclusion are not calculated appropriately. Descriptive statistics do not allow to take into account other factors that will affect the likelihood of having force used or being arrested. There are a large number of factors that go into a decision to arrest or use force. While no statistical analysis can be perfect, if we are studying racial disparities in these areas, inclusion of as many of these important variables are necessary before claims of systemic bias can possibly be supported.
However, problems in methodologies arise in attempting to show that systemic bias in policing is present as much criminal justice research is conducted using pre-existing data; the data sets used for these analyses can be lacking important variables and their associated data. Criminal justice research has already established the wide variety of variables that strongly affect police activities like arrest and use of force outside of any racial effect. Research with these data sets may not provide a complete picture of what is driving criminal justice outcomes when they lack important variables related to racial difference that affect outcome measures.
When examining police behavior for bias, the statistical model depends on the variables available within the dataset(s) used. If only a few independent controlling variables are used, the explanatory power is reduced. Many statistical models that rely on descriptive statistics, and even some multi-variate models will lack important variables that could influence racial disparities.
Some are geographical like neighborhood crime rates, neighborhood race demographics, crime distribution, and policing initiatives put in place by police or city management. Some are situational context variables at the scene like violence occurring, the presence of weapons and evidence and may be hard to measure such as the victims’ desire for charges or arrests, witness cooperation and truthfulness, degree of evidence present, and factors contributing to the risk of officer safety like physical location, surroundings, number of people present, There also some contextual variables related to the subject, variables that may be disproportionately associated with minorities, like the seriousness of the crime, racial differences in crime involvement, presence in high crime areas, and demeanor, noncompliance and resistance. All of these variables have a significant effect on whether or not an arrest is performed or force is used. Research has demonstrated once these variables are controlled for, the racial disparities shrink or disappear. The racially disproportionate amount of arrests and use of force stem from the other variables that drive crime and police activity, which negates a racial bias argument.
While comprehensive, well-constructed studies with quality data sets and rigorous methodologies do not suggest there is any systemic racial bias in policing, the search for racial bias has continued by examining behavior, specifically whether unconscious biases about race may generate racial disparities in the use of force. Even the concept of implicit bias doesn’t clearly indicate racial bias might occur in behavior. Research has shown that not everyone has the same implicit biases or even biases against outgroups, and that implicit bias can be controlled or mitigated, as well as that outside factors, like current events, situational factors, and physiological effects, can affect the level of implicit bias and how well it can be controlled.
The Implicit Association Test is a well established technique for examining biases held by individuals. It seems to also be well-established from the shoot/no shoot video simulations that not just police officers but the general population, of all races, adhere to the stereotype of the violent Black male by being quicker to react to armed Black targets and slower to react to armed White targets. However, some research actually shows officers reacting slower to armed Black targets, perhaps using an overabundance of caution that might result from implicit bias awareness training or from an awareness of the scrutiny that police/Black citizen interactions can generate. In regards to the behavioral response, while the general public tended to make significantly more errors by shooting unarmed Black targets, officers did not, further indicating a lack of racial bias.
However, none of this addresses policing activities in general. Despite the existence of implicit bias, there is no way to measure it at any specific moment in a real police interaction, and subject behavior and demeanor will all have an influence on the strength of the implicit bias, that if present, may also affect police behavior. It is unknown whether in any given instance of police contact to what degree, if any, implicit bias is at work, and what degree other factors affecting police behavior are accounted for such as, types of calls for service, suspect description, witness statements evidence of a commission of a crime, and objectively suspicious behavior (though Richardson, 2010, claimed that police cannot objectively gauge suspicious behavior because of the racial bias that must be inherently present in policing because of their contact with minority criminality) may influence or account for officer behavior.
When comprehensive modeling is used, research has shown that often the disparities between Black and White outcomes disappear or become reduced to a level that could be explained by other important variables not measured. Without a statistically significant indication of any sizable racial disparity in arrests and use of force, with indications that a small percent of officers may generate notable disparities, without clear evidence of systemic overtly racist departmental policies or behavior, and a lack of proof that implicit bias against minorities is ongoing in police interactions and actually affects their behavior, the argument of systemic racial bias in policing is not supported. While it appears that there is no evidence of systemic racial bias in policing, it doesn’t mean that improvements can’t be made in race relations with the police and that we shouldn’t continue to monitor policing for indications of racial discrimination. There are undoubtedly “bad apples” in police departments around the country who are racially biased and who have very profound effects on peoples’ of color lives. But in a desire to eliminate racism from policing, efforts to identify it must be conducted with scientific rigor, because the truth matters. As Fryer (2018) said, “Of course, Black lives matter as much as any other lives. Yet we do this principle a disservice if we do not adhere to strict standards of evidence and take at face value descriptive statistics that are consistent with our preconceived ideas. “Stay woke”-but critically so”.
REFERENCES
Arndt, M., Stolzenberg, L., & D’Alessio, S. J. (2020). The effects of race and physical evidence on the likelihood of arrest for homicide. Race and justice, 2153368719900358.
Atiba Goff, P., & Barsamian Kahn, K. (2012). Racial bias in policing: Why we know less than we should. Social Issues and Policy Review, 6(1), 177-210.
Balko, R. (2018). There’s overwhelming evidence that the criminal-justice system is racist. Here’s the proof. The Washington Post, 18.
Briggs, S., & Opsal, T. (2012). The influence of victim ethnicity on arrest in violent crimes. Criminal Justice Studies, 25(2), 177-189.
Brown, R. A. (2005). Black, white, and unequal: Examining situational determinants of arrest decisions from police–suspect encounters. Criminal Justice Studies, 18(1), 51-68.
Brown, R. A., & Frank, J. (2006). Race and officer decision making: Examining differences in arrest outcomes between black and white officers. Justice quarterly, 23(1), 96-126.
Buehler, J. W. (2017). Racial/ethnic disparities in the use of lethal force by US police, 2010–2014. American journal of public health, 107(2), 295-297.
Camplain, R., Camplain, C., Trotter, R. T., Pro, G., Sabo, S., Eaves, E., … & Baldwin, J. A. (2020). Racial/ethnic differences in drug-and alcohol-related arrest outcomes in a southwest county from 2009 to 2018. American journal of public health, 110(S1), S85-S92.
Cesario, J., Johnson, D. J., & Terrill, W. (2019). Is there evidence of racial disparity in police use of deadly force? Analyses of officer-involved fatal shootings in 2015–2016. Social psychological and personality science, 10(5), 586-595.
Cooley, E., & Brown-Iannuzzi, J. (2019). Liberals perceive more racism than conservatives when police shoot Black men—But, reading about White privilege increases perceived racism, and shifts attributions of guilt, regardless of political ideology. Journal of Experimental Social Psychology, 85, 103885.
Correll, J., Park, B., Judd, C. M., & Wittenbrink, B. (2002). The police officer’s dilemma: Using ethnicity to disambiguate potentially threatening individuals. Journal of personality and social psychology, 83(6), 1314.
Correll, J., Park, B., Judd, C. M., Wittenbrink, B., Sadler, M. S., & Keesee, T. (2007). Across the thin blue line: police officers and racial bias in the decision to shoot. Journal of personality and social psychology, 92(6), 1006.
Correll, J., & Keesee, T. (2009). Racial bias in the decision to shoot. The Police Chief, 76(5), 54-57.
Correll, J., Wittenbrink, B., Park, B., Judd, C. M., & Goyle, A. (2011). Dangerous enough: Moderating racial bias with contextual threat cues. Journal of experimental social psychology, 47(1), 184-189.
Correll, J., Hudson, S. M., Guillermo, S., & Ma, D. S. (2014). The police officer’s dilemma: A decade of research on racial bias in the decision to shoot. Social and Personality Psychology Compass, 8(5), 201-213.
D’Alessio, S. J., & Stolzenberg, L. (2003). Race and the probability of arrest. Social forces, 81(4), 1381-1397.
Drakulich, K., Wozniak, K. H., Hagan, J., & Johnson, D. (2020). Race and policing in the 2016 presidential election: Black lives matter, the police, and dog whistle politics. Criminology, 58(2), 370-402.
Drawve, G., Thomas, S. A., & Walker, J. T. (2014). The likelihood of arrest: A routine activity theory approach. American Journal of Criminal Justice, 39(3), 450-470.
Durlauf, S. N., & Heckman, J. J. (2020). An Empirical Analysis of Racial Differences in Police Use of Force: A Comment. Journal of Political Economy, 128(10), 3998-4002.
Elliott, D. S. (1995). Lies, damn lies, and arrest statistics. Boulder, CO: Center for the Study and Prevention of Violence.
Engel, R. S., Sobol, J. J., & Worden, R. E. (2000). Further exploration of the demeanor hypothesis: The interaction effects of suspects’ characteristics and demeanor on police behavior. Justice quarterly, 17(2), 235-258.
Evans, D. N., Maragh, C. L., & Porter, J. R. (2014). What Do We Know About NYC’s Stop and Frisk Program?: A Spatial and Statistical Analysis. Advances in Social Sciences Research Journal, 1(2), 130-144.
Fielding-Miller, R., Cooper, H. L., Caslin, S., & Raj, A. (2020). The interaction of race and gender as a significant driver of racial arrest disparities for African American men. Journal of urban health, 97(1), 112-122.
Finkeldey, J. G., & Demuth, S. (2019). Race/ethnicity, perceived skin color, and the likelihood of adult arrest. Race and Justice, 2153368719826269.
Fridell, L. A. (2016). Racial aspects of police shootings: Reducing both bias and counter bias. Criminology & Pub. Pol’y, 15, 481.
Fridell, L., & Lim, H. (2016). Assessing the racial aspects of police force using the implicit-and counter-bias perspectives. Journal of Criminal Justice, 44, 36-48.
Fryer Jr, R. G. (2018, May). Reconciling results on racial differences in police shootings. In AEA papers and proceedings (Vol. 108, pp. 228-33).
Fryer Jr, R. G. (2019). An empirical analysis of racial differences in police use of force. Journal of Political Economy, 127(3), 1210-1261.
Futterman, C. B., Mather, H. M., & Miles, M. (2007). The Use of Statistical Evidence to Address Police Supervisory and Disciplinary Practices: The Chicago Police Department’s Broken System. DePaul J. Soc. Just., 1, 251.
Gelman, A., Fagan, J., & Kiss, A. (2007). An analysis of the New York City police department’s “stop-and-frisk” policy in the context of claims of racial bias. Journal of the American statistical association, 102(479), 813-823.
Gibbons, F. X., Fleischli, M. E., Gerrard, M., Simons, R. L., Weng, C. Y., & Gibson, L. P. (2020). The impact of early racial discrimination on illegal behavior, arrest, and incarceration among African Americans. American Psychologist, 75(7), 952.
Hall, A. V., Hall, E. V., & Perry, J. L. (2016). Black and blue: Exploring racial bias and law enforcement in the killings of unarmed black male civilians. American Psychologist, 71(3), 175.
Hehman, E., Flake, J. K., & Calanchini, J. (2018). Disproportionate use of lethal force in policing is associated with regional racial biases of residents. Social psychological and personality science, 9(4), 393-401.
Hollinger, R. C. (1984). Race, occupational status, and pro-active police arrest for drinking and driving. Journal of Criminal Justice, 12(2), 173-183.
Hollis, M. E., & Jennings, W. G. (2018). Racial disparities in police use-of-force: a state-of-the-art review. Policing: An International Journal.
Holroyd, J. (2015). Implicit racial bias and the anatomy of institutional racism. Criminal Justice Matters, 101(1), 30-32.
James, L., Klinger, D., & Vila, B. (2014). Racial and ethnic bias in decisions to shoot seen through a stronger lens: Experimental results from high-fidelity laboratory simulations. Journal of Experimental Criminology, 10(3), 323-340.
James, L. (2018). The stability of implicit racial bias in police officers. Police Quarterly, 21(1), 30-52.
Johnson, D. J., & Wilson, J. P. (2019). Racial bias in perceptions of size and strength: The impact of stereotypes and group differences. Psychological science, 30(4), 553-562.
Kahn, K. B., Goff, P. A., Lee, J. K., & Motamed, D. (2016). Protecting whiteness: White phenotypic racial stereotypicality reduces police use of force. Social Psychological and Personality Science, 7(5), 403-411.
Kirk, D. S. (2008). The neighborhood context of racial and ethnic disparities in arrest. Demography, 45(1), 55-77.
Kochel, T. R., Wilson, D. B., & Mastrofski, S. D. (2011). Effect of suspect race on officers’ arrest decisions. Criminology, 49(2), 473-512.
Lantz, B., & Wenger, M. R. (2019). The co-offender as counterfactual: A quasi-experimental within-partnership approach to the examination of the relationship between race and arrest. Journal of experimental criminology, 1-24.
Lautenschlager, R., & Omori, M. (2019). Racial threat, social (dis) organization, and the ecology of police: Towards a macro-level understanding of police use-of-force in communities of color. Justice quarterly, 36(6), 1050-1071.
Legewie, J. (2016). Racial profiling and use of force in police stops: How local events trigger periods of increased discrimination. American journal of sociology, 122(2), 379-424.
Levin, J., & Thomas, A. R. (1997). Experimentally manipulating race: Perceptions of police brutality in an arrest: A research note. Justice Quarterly, 14(3), 577-586.
Levinson, J. D., & Young, D. (2009). Different shades of bias: Skin tone, implicit racial bias, and judgments of ambiguous evidence. W. Va. L. Rev., 112, 307.
Lum, C. (2011). The influence of places on police decision pathways: From call for service to arrest. Justice Quarterly, 28(4), 631-665.
Lytle, D. J. (2014). The effects of suspect characteristics on arrest: A meta-analysis. Journal of Criminal Justice, 42(6), 589-597.
Ma, D. S., Correll, J., Wittenbrink, B., Bar-Anan, Y., Sriram, N., & Nosek, B. A. (2013). When fatigue turns deadly: The association between fatigue and racial bias in the decision to shoot. Basic and applied social psychology, 35(6), 515-524.
Mears, D. P., Craig, M. O., Stewart, E. A., & Warren, P. Y. (2017). Thinking fast, not slow: How cognitive biases may contribute to racial disparities in the use of force in police-citizen encounters. Journal of Criminal Justice, 53, 12-24.
Miller, S. L., Zielaskowski, K., & Plant, E. A. (2012). The basis of shooter biases: Beyond cultural stereotypes. Personality and Social Psychology Bulletin, 38(10), 1358-1366.
Mitchell, O., & Caudy, M. S. (2015). Examining racial disparities in drug arrests. Justice Quarterly, 32(2), 288-313.
Morrow, W. J., White, M. D., & Fradella, H. F. (2017). After the stop: Exploring the racial/ethnic disparities in police use of force during Terry stops. Police Quarterly, 20(4), 367-396.
Nieuwenhuys, A., Savelsbergh, G. J., & Oudejans, R. R. (2012). Shoot or don’t shoot? Why police officers are more inclined to shoot when they are anxious. Emotion, 12(4), 827.
Penner, A. M., & Saperstein, A. (2015). Disentangling the effects of racial self-identification and classification by others: the case of arrest. Demography, 52(3), 1017-1024.
Petersilia, J. (1983). Racial disparities in the criminal justice system (Vol. 2947). Santa Monica, CA: Rand Corporation.
Plant, E. A., Goplen, J., & Kunstman, J. W. (2011). Selective responses to threat: The roles of race and gender in decisions to shoot. Personality and Social Psychology Bulletin, 37(9), 1274-1281.
Pratt-Harris, N. C., Sinclair, M. M., Bragg, C. B., Williams, N. R., Ture, K. N., Smith, B. D., … & Brown, L. (2016). Police-involved homicide of unarmed Black males: Observations by Black scholars in the midst of the April 2015 Baltimore uprising. Journal of Human Behavior in the Social Environment, 26(3-4), 377-389.
Raphael, S., & Rozo, S. V. (2019). Racial disparities in the acquisition of juvenile arrest records. Journal of Labor Economics, 37(S1), S125-S159.
Ramchand, R., Pacula, R. L., & Iguchi, M. Y. (2006). Racial differences in marijuana-users’ risk of arrest in the United States. Drug and alcohol dependence, 84(3), 264-272.
Reinka, M. A., & Leach, C. W. (2017). Race and reaction: Divergent views of police violence and protest against. Journal of Social Issues, 73(4), 768-788.
Richardson, L. S. (2010). Arrest Efficiency and the Fourth Amendment. Minn. L. Rev., 95, 2035.
Ridgeway, G. (2007). Analysis of racial disparities in the New York Police Department’s stop, question, and frisk practices. Rand Corporation.
Ridgeway, G., & MacDonald, J. M. (2009). Doubly robust internal benchmarking and false discovery rates for detecting racial bias in police stops. Journal of the American Statistical Association, 104(486), 661-668.
Rivera, M. A., & Ward, J. D. (2017). Toward an analytical framework for the study of race and police violence. Public administration review, 77(2), 242-250.
Roberts, D. E., & Race, V. (1999). the Social Meaning of Order-Maintenance Policing, 89 J. Crim. L. & Criminology, 775, 799-818.
Rose, J. B. G. (2018). Racial Character Evidence in Police Killing Cases. Wis. L. Rev., 369.
Ross, C. T. (2015). A multi-level Bayesian analysis of racial bias in police shootings at the county-level in the United States, 2011–2014. PloS one, 10(11), e0141854.
Ross, C. T., Winterhalder, B., & McElreath, R. (2020). Racial disparities in police use of deadly force against unarmed individuals persist after appropriately benchmarking shooting data on violent crime rates. Social Psychological and Personality Science, 1948550620916071.
Roussell, A., Henne, K., Glover, K. S., & Willits, D. (2017). Impossibility of a “reverse racism” effect: a rejoinder to James, James, and Vila. Criminal Public Policy. 2017.
Sadler, M. S., Correll, J., Park, B., & Judd, C. M. (2012). The world is not black and white: Racial bias in the decision to shoot in a multiethnic context. Journal of Social Issues, 68(2), 286-313.
Schleiden, C., Soloski, K. L., Milstead, K., & Rhynehart, A. (2020). Racial disparities in arrests: a race specific model explaining arrest rates across black and white young adults. Child and adolescent social work journal, 37(1), 1-14.
Sealock, M. D., & Simpson, S. S. (1998). Unraveling bias in arrest decisions: The role of juvenile offender type-scripts. Justice Quarterly, 15(3), 427-457.
Senholzi, K. B., Depue, B. E., Correll, J., Banich, M. T., & Ito, T. A. (2015). Brain activation underlying threat detection to targets of different races. Social neuroscience, 10(6), 651-662.
Tapia, M. (2011). Gang membership and race as risk factors for juvenile arrest. Journal of Research in Crime and Delinquency, 48(3), 364-395.
Tobias, H., & Joseph, A. (2020). Sustaining systemic racism through psychological gaslighting: Denials of racial profiling and justifications of carding by police utilizing local news media. Race and Justice, 10(4), 424-455.
Tolliver, W. F., Hadden, B. R., Snowden, F., & Brown-Manning, R. (2016). Police killings of unarmed Black people: Centering race and racism in human behavior and the social environment content. Journal of Human Behavior in the Social Environment, 26(3-4), 279-286.
Weisburd, D., Wooditch, A., Weisburd, S., & Yang, S. M. (2016). Do stop, question, and frisk practices deter crime? Evidence at microunits of space and time. Criminology & public policy, 15(1), 31-56.
Welsh, M., Chanin, J., & Henry, S. (2020). Complex Colorblindness in Police Processes and Practices. Social Problems.
Wilson, J. P., Hugenberg, K., & Rule, N. O. (2017). Racial bias in judgments of physical size and formidability: From size to threat. Journal of personality and social psychology, 113(1), 59.
