Segregated Spaces and Separated Races: The Relationship Between State-Sanctioned Violence, Place, and Black Identity

This article explores segregation and the social status of black people, focusing in particular on the ripple effects of Plessy v. Ferguson on policing in the United States. Specifically, we ask how the legacy of Plessy v. Ferguson has helped maintain state-sanctioned racially based violence. We draw from Mapping Police Violence, which compiles data on the number of police-involved homicides in large police departments in the United States from 2013 to 2017. Using these data, we analyze the relationship between space and the number of deaths of black people caused by police. The findings reveal a positive association between segregated communities, the percentage of black people in a population, and the number of police killings of black people. This study provides a unique opportunity to examine the impact that segregation and place have on state-sanctioned violence, particularly for those who identify or are identified as black.

The history of the United States has been tainted by the social construction of race and reflects the calamities and conquests of society's tribulations. Rhetorically, America was founded on the principle that all men are created equal, thereby placing value on individual rights and freedoms. In practice, however, the nation began as a slaveholding society (Williams et al. 2016). Consequently, many Africans in America were, in the words of Jefferson Davis, "stamped from the beginning" (Kendi 2016). This conundrum resulted in an American dilemma and a lasting legacy of socially constructed and politically reinforced practices that limited access to space along racial lines (Myrdal 1944).

Segregation was "a brutal regime of systematic degradation and oppression that was all encompassing in the South and widely practiced in the North" long before the Plessy v. Ferguson Supreme Court ruling in 1896 (Ware 2018, vii).¹ The Plessy decision, however, did far more than simply uphold segregation. The ruling doubled down on the constitutionality of "separate but equal," advancing racial segregation as public policy and formalizing a race-based caste system in the United States. With the Plessy decision, the system of Jim Crow thrived without legal scrutiny and allowed for racial segregation to persist while setting the course for long-standing residential segregation across the country (Alexander 2012).

The racial caste system maintained by Plessy and strengthened with Jim Crow fostered the perfect political and social conditions for the continuance of residential segregation. Even after the passage of civil rights acts in the nineteenth and twentieth centuries, African Americans and black people continued to overwhelmingly be relegated to the worst parts of town (Alexander 2012). Further, Plessy "confirmed state power to assign personal identity in classes with consequences of inclusion and exclusion" (Davis 2021, this issue). Thus Plessy's ever-present legacy is shaped by the long-term effects of segregation on the social status of black people and African Americans.

The Plessy effect, according to Lorenzo Morris (1979), offers the guise of equality in broad and elusive ways but avoids the actual implementation of policy and practices that provide tangible and meaningful equality for black people. From the blatant discrimination in the case's decision to the notion of colorblindness in the dissent, Plessy has set a precedent allowing for public agencies and actors to use racial hierarchy as a tool to explicitly and implicitly segregate public spaces. Plessy's ripple effect is felt across time and in all aspects of life, including but not limited to inequitable access to quality public and higher education (Henry 1998), public housing (Peel, Pickett, and Buehl 1970) and the criminalization and incarceration of black people (Brewer and Heitzeg 2008). Ultimately, Plessy's legacy reinforces difference as deviance, deeming black people unfit and thus socially illegitimate (Hutchison 2011). The Plessy effect advances Darwinistic thought of blackness as inferior and thus requiring and deserving of administrative and policy burdens.

The unspoken "rules" of residential segregation dictated one's worthiness and sent strong messages of who belonged where, hoarding opportunity "and asset accumulation for whites while confining aggrieved communities of color to impoverished, under-resourced and criminogenic neighborhoods" (Lipsitz 2015, 119). Consequently, the negative social constructions of criminality that were linked to blackness also became associated with neighborhoods deemed appropriate for black and African American residence. The result was the overpolicing of such communities and the continuance of social control of black and brown bodies.

The Plessy decision left a legacy that defines blackness in stark opposition to whiteness—"a status conferring distinct—yet conjoined—social, political, and economic freedoms across a vertiginously unequal property order" (Singh 2014, 1019). Whereas whiteness "suggested a relationship between differential human valuation and materially valuable access" (Singh 2014, 1019), blackness signified a "legally sanctioned inferior status" (Alexander 2012, 195). The negative status conferred on blackness affected the environments where black people lived and resulted in the underpolicing or tolerating of certain crimes in black neighborhoods while simultaneously overpolicing these same communities through the use of excessive force and violence against residents (Hawkins and Thomas 2013). This duality crafted a context in which state-sanctioned racially based violence is not only condoned but also led by police officials and agencies.

This article explores the long-term effects of segregation on black and African American people in the United States. Although we do not seek to make direct connections between the legal concept of "separate but equal" on train cars in Louisiana, as Justice Henry Billings Brown wrote for the majority in the Plessy decision, we do offer that the ripple effect of whiteness as the dominant race—as maintained by the Supreme Court's decision and thus its legacy—has shaped today's institutions including but not limited to housing and policing. Specifically, we examine the relationship between the legacy of Plessy v. Ferguson and policing institutions in the United States by analyzing state-sanctioned violence perpetrated against black people. Bearing on the notions of whiteness and blackness in the context of space (Alexander 2012; Singh 2014), we distinguish between the concepts of "black space"—a "destitute and fearsome locality"—and "white space"—predominantly white and off-limit spaces for black people—as presented by Elijah Anderson (2015, 10). Our proposition is that the creation and perpetuation of black space with the "second-class status" construction of blackness (Ware 2021, this issue) has resulted in a disproportionate number of black people killed by law enforcement actors.

To test this proposition, we drew data from Mapping Police Violence, an advocacy organization that collected data on the number of known police killings across the United States. We used a sample of the number of killings in the hundred largest U.S. cities by population for the years 2013 to 2017. We used these data, coupled with additional demographic and socioeconomic factors, to analyze the factors associated with an increase in the number of black people killed by police. Through these analyses, we seek to understand the role of race and place in the number of police homicides of black people. Overall, this study provides a unique opportunity to draw correlations between space, policing, race, and state-sanctioned violence.

DESEGREGATED, YET STILL SEPARATE

In 1917, the United States Supreme Court ended legalized residential segregation with Buchanan v. Warley (Rice 1968, 179).² Yet the legacy of the Plessy ruling, rendered merely twenty years earlier, was long lasting and far reaching, aiding segregation to endure as the customary practice in states across the country. Just nine years after the Buchanan decision, the Supreme Court authorized the use of restrictive covenants in Washington, D.C., with Corrigan v. Buckley.³ Residential segregation ordinances were passed in northern states such as Maryland, racially restrictive covenants were deployed on a wide scale by the real estate industry, and redlining practices associated property values with the racial composition of neighborhoods (Ware 2021, this issue). In 1948, although not overturning Corrigan, the Shelley v. Kraemer decision determined racially restrictive covenants unenforceable and Jones v. Mayer Co. (1968) barred racial discrimination in both public and private sales or rentals of property.⁴ Although seemingly advancing housing equality, the Supreme Court also made it more difficult to prove housing discrimination. In Village of Arlington Heights v. Metropolitan Housing Development Corp. (1977), the Court determined that racially discriminatory intent was required to bring such a claim and imposed major limitations on the application of disparate impact in Texas Department of Housing and Community Affairs v. The Inclusive Communities Project, Inc. (2015).⁵

Narratives advanced after World War II conceived of segregation as comprising low-income "urban" centers and affluent suburban rings. Contemporary residential segregation, however, has evolved to become more complex. William Goldsmith and Edward Blakely argue that this complexity is akin to a checkerboard of "prosperity and poverty" (2010, 112). These complex checkerboard spaces suggest that though the city center and suburban rings may be composed of diverse social groups and income strata, sharp geographic variations by neighborhood in income, race, ethnicity, and advantage still distinguish daily opportunities and experiences. In this regard, segregation is at the level where people live their daily lives. Even in seemingly diverse communities such as San Francisco and New York City, variations in neighborhood wealth create economically, educationally, and racially segregated neighborhoods.

The intersection of race, economic status, and educational attainment in communities across the United States operates to either promote or limit opportunities for mobility and residential choice. The more affluent can choose a housing location based on personal preferences, neighborhood amenities, and accessibility. Those with lower incomes, however, have less choice and are limited to passively moving out of declining neighborhoods as opposed to actively choosing neighborhoods to move into based on individual or familial preferences (Goldsmith and Blakely 2010). For many, the existence of segregated neighborhoods is not a surprising phenomenon. This is evidenced by the fact that the affluent flee declining neighborhoods and those of lower economic status remain in areas that are economically shrinking or stagnant as well as characterized by underfunded schools and environmental pollutants (Goldsmith and Blakely 2010).

The variability in U.S. community characteristics is considerable and the customary segregation associated with such variation is the embodiment of what Manning Marable (2000) calls the "underdeveloped black America." Marable argues that the economic and political history of the United States has led to the systematic underdevelopment of black people. Specifically, segregation, caused by a racial caste system, has resulted in unemployment, underemployment, economic exploitation, and political disenfranchisement. Michelle Alexander (2012) adds that racial segregation is what relegated black people to the worst communities, creating fertile ground for the underdevelopment of black neighborhoods and people across the country.⁶

Through legal segregation advanced by the Plessy ruling, these communities received far fewer governmental resources and services than white communities, and the black residents living in these neighborhoods were seen as inferior (Alexander 2012). For black people in the United States, the notion of "separate but equal" was nonexistent. Legal and customary segregation expanded racial stigmata in the United States. The notion of blackness and the communities where black people lived became synonymous with dilapidation, poverty, and crime. Such racial stigmata persist today and operate to maintain customary segregation along racial, economic, and residential lines.

Stereotypes of unworthiness spurred by economic, racial, and residential segregation have also operated to conflate race and crime, creating seemingly natural associations between the two. The everyday lived experiences of those living in lower-income communities are rarely seen in plain sight. In fact, customary segregation isolates those living in lower-income communities from their more affluent neighbors and vice versa. Such conflations of race and crime have led to associations of wrongdoing and criminality that are almost exclusively linked to racial identity and economic status, yet understood as truth. Notions of increased crime and criminal proclivity combined with the directive to control illegal activity have resulted in the overpolicing of communities perceived to be associated with higher crime rates (Weitzer, Tuch, and Skogan 2008). As a result, individuals living in overwhelmingly black and low-income communities find themselves interacting with police at much greater rates than those who can make active residential decisions.

RACE, SPACE, AND SOCIAL CONTROL: THE REALM OF LAW ENFORCEMENT

Since the founding of the United States, two criminal justice systems existed—one for black people and one for white people. Enslaved Africans and later black Americans were forced to navigate both formal law enforcement, by way of policing organizations, and the informal terroristic social control practices of the Ku Klux Klan and lynching mobs (Ware 2018). Those (suspected of) committing a crime or engaged in behavior considered unacceptable for black people were often forced to deal with these systems. Operating often in tandem, these formal and informal enforcement practices embodied violence and intimidation as social control tactics designed to keep black people in their metaphorical and physical spaces. Lynching is "an act of homicidal aggression committed by one people against another through mob action for the purpose of suppressing either some tendency in the latter to rise from an accommodated position of subordination or for subjugating them further to some lower social status" (Cox 1945, 576). In the United States, the act of lynching served as an attack against the entire black community rather than an individual, in that, "the destruction of almost any Negro will serve the purpose as well as that of some particular one" and operated as "the most powerful and convincing form of racial repression" (Cox 1945, 581). Formal law enforcement agencies as well as local and state judicial systems offered limited protections against lynching and lynch mobs, reifying to black people that the criminal legal system in the United States was not designed to protect their rights or to apply the law in their favor (Cox 1945). Whites, in general, were the beneficiaries of the system. Yet in the context of the labor-intensive economy in the South that depended on black workers, limited protection was afforded in the early 1900s to black people in some instances to stem the possible out-migration of African Americans to the North (Beck and Tolnay 1990; Tolnay and Beck 1995). Nonetheless, the two-tiered justice system also operated to reinforce the tension between black communities and law enforcement agencies. Tensions between the police and racial-ethnic minority communities is a persistent problem across the globe. This has created what has been described as "nervous landscapes," where spatial separation between races has been legislated and implemented (Byrne 2003a). Consequently, segregation has emerged as the spatial face of racism (Byrne 2003b) and this policy has been enforced by formal (institutional or organizational) and informal (individual and collective or communal) approaches to social control.

Similar to what occurs in other settings across the globe, the current state of police-community relations in the United States is in a perilous position. Recent, highly publicized encounters between police and black people have exacerbated already low levels of public trust and confidence in local law enforcement. Tensions between the police and the black community have been and continue to be one of the most pressing issues facing U.S. police organizations (Williams 1998; Barak et al. 2001; Webs-dale 2001; Culver 2004; Hawkins and Thomas 2013). Such tensions arise from the interactive relationship between race, place, and policing (Bass 2001). As true of residential segregation, the history of policing in the United States has been tainted by legally sanctioned, disparate service delivery and tarnished by the enforcement of racially motivated laws and statutes. The present-day relationship between black communities and the police is one affected by the historical legacies of slavery, segregation, discrimination, and mass incarceration, all part and parcel of racism at the formal and informal levels (Russell 1998; Barak et al. 2001; Alexander 2012; Rothstein 2017; Owusu-Bempah 2017). The antecedent of these policies and practices has been the social construction of race.

Race is a historical social construction yet it continues to be a significant feature of U.S. society, especially because it reinforces and fortifies the lasting, stigmatized legacy of blackness (West 1994). Glenn Loury (2002) uses three dimensions to emphasize the historical germaneness of race and surmises the impact it has in contemporary society. The first is the relative ease in identifying the other, the second is the relative unchanging quality of this construct, and the third is the social significance of this designation. Loury's concept of race is a byproduct of social construction, yet bolsters an additional construct of stigma. Stigma has been described as a bodily sign that exposes something unusual and negative about the moral status of the signifier (Goffman 1963).

The coupling of race and stigma—in terms of blackness—resulted in an insidious theory of black criminality that has transcended continental boundaries and time and has been embraced by secular and sacred institutions (Gabbidon 2015; Williams et al. 2016). From the writings of John Van Evrie (1861), the U.S. Supreme Court (1860), Cesare Lombroso (2006), and others, a notion that blacks were biologically inferior and innately criminalistic advanced. Hence they advocated for the mass incarceration and enhanced criminal penalties for black people, a practice echoed by U.S. public institutions as part of the country's war on drugs (Alexander 2012). These punitive measures are part and parcel of state-sanctioned violence.

The stigma that accompanied blackness as a race in the socially constructed and politically reinforced global and U.S. society of yesterday, resulted in Billy Close's (1997) conception of black crimmythology—a biased social cognition that continues to affect Americans of African descent. Black crimmythology is used to describe the numerous faulty myths and pseudo-scientific facts, historically and contemporarily, which have been generated by racist and/or misguided individuals to explain black criminality. It projects blackness as a defect best understood as a manifestation of an immutable, transgenerational, and permanent condition of black inferiority (Close 1997).

These historical notions of blackness, black criminality, and black crimmythology reinforce and propel into the future Howard Becker's "different as deviant narrative" (1963), while continuing to color in shades of black and brown the portrait of America's symbolic assailant (Skolnick 1998). This lasting legacy of black ness projects all of the assumed dangers and criminality of what Elijah Anderson (2012) has described as the "iconic ghetto"—an individual black male who symbolizes the violent, impoverished, crime-prone, and drug-infested environment of black people. This ascription of ghetto manifested in the black body continues to persist as U.S. society evolves away from the segregated black and white spaces of our past and into the relatively more cosmopolitan or, seemingly, racially diverse spaces or enclaves of our present and projected future (Anderson 2011, 2012, 2015).

Franklin Zimring (2017) conducted a comprehensive study of police use of lethal force in the United States. Simultaneously, he explored the phenomena of violence by police against civilians and violence by civilians toward police. His findings reveal that of the 1,100 killings by police in the United States in 2015, the death rates for African American and Native Americans were twice their respective shares of the population and, among other things, when police officers felt threatened or attacked, they were more likely to fatally shoot a civilian. This was particularly the case when officers were alone and feared that a gun was involved. Zimring's research calls into question the theoretical civility associated with our contemporary times and our cosmopolitan canopies. In particular, his findings seem to reinforce the notion that blackness endures as a powerful source of stereotype, prejudice, and discrimination that impacts the American psyche. Today's reminders seem ever present—the deaths of Breonna Taylor, Ahmaud Arbery, George Floyd, and the all-too-often social media postings where black people are assumed dangerous as they engage in mundane tasks (exercising, bird-watching, golfing)—of the threat and promise that random and arbitrary violence can occur when black people venture beyond the traditional "black space" into the "white space" and the cosmopolitan canopy (Anderson 2011, 2015). This brings into focus a poignant observation of Loury: the need to continue exploring "the enigma of the stigma" and not necessarily "the figment of the pigment" (2002, 143). Moreover, this sets the stage for our inquiry.

METHODOLOGY

This study examines how the creation and maintenance of black space with second-class status—one aspect of Plessy's legacy—has led to a disproportionate number of black people killed by police in predominantly black communities. We operationalize state-sanctioned violence as the actions of U.S. law enforcement actors that result in a disproportionately high rate of deaths of black people. Based on a literature review, we argue that black space is associated with a higher rate of police homicides of black people. Primary data are derived from Mapping Police Violence, an advocacy organization that has collected data on the number of known police killings across the country. Although Mapping Police Violence currently provides data for the years between 2013 and 2019, the 2017 data were the most recent available to us at the time of this study. We used a sample of the number of killings in the hundred largest U.S. cities by population, ranging from around two hundred thousand to about eight million residents. These data are limited to cities, not to metropolitan areas or a larger geographic unit. According to the organization, a police-involved homicide refers to a case in which an on-duty or off-duty police officer either intentionally or accidentally kills a person in the process of chasing, beating, arresting, restraining, shooting, pepper spraying, Tasering, or harming the person by other use-offorce actions (Mapping Police Violence 2019). Each homicide is coded by the police department and the gender and race of the victim. The dataset was assembled based on data from the three largest, most comprehensive crowd-sourced databases on police killings in the country: FatalEncounters.org, the U.S. Police Shootings Database, and KilledbyPolice.net. The majority of the data are derived from KilledbyPolice.net because, according to Mapping Police Violence, this was fact-checked by Nate Silver's FiveThirtyEight research organization. KilledByPolice.net aggregates links from major social media sites to news articles on police-related killings, and FiveThirtyEight selected a random sample of approximately 10 percent of the links from the original news stories (Fischer-Baum and Johri 2014). Reuben Fischer-Baum and Al Johri find that all of the links were accurate and that 93 percent of reported incidents were either incidents in which the victims were fatally shot by on-duty officers or the victims died after being restrained by force in some way by law enforcement.

Mapping Police Violence undertook extensive data gathering of its own to improve the accuracy and completeness of the data from the three crowdsourced databases on police killings in the United States. Using cases from FatalEncounters.org, the U.S. Police Shootings Database, and KilledbyPolice.net as the starting point, the organization used social media, newspaper accounts, obituaries, police reports, and other sources to confirm the occurrence of a death and identify the race of approximately 90 percent of victims. In addition, it used other crowdsourced databases, media sources, and criminal records databases, and performed additional inquiries to fill in additional information, such as whether the victim was armed or unarmed. While some police killings may not have been reported, such as those unreported by the media, Mapping Police Violence estimates that their figures capture approximately 93 percent of all police killings since 2013.

We supplement the analysis with additional data derived from the U.S. Census Bureau American Community Survey and FBI Uniform Crime Reports from 2013 to 2017.

Variables

The dependent variable is a count of the number of police homicide victims who were black or African American for each year between 2013 and 2017. Count data follow a Poisson distribution, and the descriptive statistics and a significant likelihood ratio test suggest that a Poisson binomial model is the best estimator. To account for unequal exposure to the number of black homicides by police in each city, we account for the total number of homicides by each police department as the exposure variable in the model.

We include several independent variables that represent "black space." The first measure is a binary indicator coded 1 if a state is a former Confederate state (yes=1; no=0). Although the "separate but equal" doctrine applied to all states, and "white supremacy was institutionalized and crystallized all over the country," more pervasive and blatant instances of segregation and discrimination were perpetuated and documented throughout southern states that coincide with the former Confederacy (powell, Myers, and Gooden 2021, this issue).

Cities with a large population of black people may be associated with a higher incidence of police violence because officers often engage in social control policing, particularly toward groups considered to be threatening the social order (Carbado 2017). We include the percentage of black population in a city to capture this expectation. Based on data availability, we also include two specific measures of residential segregation (Massey and Denton 1988). First, the black-white dissimilarity index is a measure of segregation based on the relative separation or integration of groups, specifically within the context of the black population, across all neighborhoods of a city. The second is the spatial proximity index, which measures the average of intragroup proximities for the minority and majority populations, weighted by the proportions that each group represents of the total population. For both measures, a higher score represents higher dissimilarity and more racial clustering among neighborhoods.

To improve model specification, we include the Gini index, which is a measure of income inequality operationalized as a function of the cumulative share of people from lower income to the cumulative share of income earned. A perfect 0 corresponds to perfect income equality, whereas 1 corresponds to perfect income inequality. Additionally, we include a poverty rate variable because residential segregation has led to the concentration of poverty in many communities of color (Osypuk et al. 2009). If police violence serves as a method of social control for groups that threaten the social order, communities with higher poverty rates may experience greater violence. We derive data for these variables from the U.S. Census Bureau American Community Survey for each year between 2013 and 2017.

We also include a variable for the percentage of black officers in the police department involved in the police killing. We derive this variable from the 2013 Law Enforcement Management and Administrative Statistics survey, the most recent version at the time of this study, which contains information regarding the number of full-time sworn personnel according to race. According to representative bureaucracy theory, police organizations with a higher proportion of black officers may be more prone to advocate for the interests of black citizens and to address negative behaviors such as racial profiling and use of force against minorities (Bradbury and Kellough 2011). In this regard, we argue that more black officers may represent broader efforts to create safer spaces for black people.

We control for the violent crime and the murder rate based on the assumption that police violence is a response to the level of conflict or violence that police officers face in the environment (Nicholson-Crotty, Nicholson-Crotty, and Fernandez 2017; Ouimet and Pare 2003; Weisburd et al. 2010). We also control for the population size to account for differences in jurisdiction sizes. Other studies use population to control for the fact that larger jurisdictions are more likely to face higher crime and officers face more difficulty in addressing crime problems (Felson and Eckert 2016). For this study, however, we do not propose any particular hypothesis regarding population because our sample is derived from the hundred largest U.S. cities by population. Finally, we include year dummy variables, using 2013 as the base year, to control for unobserved year-specific effects in the data.

Estimation Method

Our data consist of a panel dataset for the years 2013 to 2017. Most of the independent variables change from one year to another, but the binary variable former Confederate state (yes=1; no=0) is a time-invariant state-level characteristic. Also, the variable for the percentage of black officers is time variant, but because of data limitations data are only available for the year 2013. This renders it difficult to estimate the model using either a fixed-effects or a random-effects model. Thus we use a mixed-effects Poisson regression, which allows us to specify former Confederate state as a random intercept while modeling the percentage of black officers as fixed effects (Hsiao 2014). However, we acknowledge the limitation of the latter because we are extrapolating trends from a single year 2013 to the

Table 1.

Descriptive Statistics

dependent variable, which spans 2013 to 2017. Therefore, we interpret the estimate for the percentage of black officers with caution because the coefficient may be biased. Table 1 presents the descriptive statistics on variable measurement, means, standard deviations, and scale ranges. Although it is not used in the main analysis, we include information on the number of citizens killed for other racial categories. Table 2 displays the correlation matrix for the variables.

FINDINGS

Before discussing the findings of the mixed-effects Poisson regression, we first present information regarding the number of persons killed by police relative to the total population. According to table 3, although the black population makes up only about 12.3 percent of the total population of the United States, the percentage of black homicides by police amounts to nearly 25 percent of all persons killed by police between 2013 and 2017. The numbers illustrate how the killing of black and African Amer ican people are disproportionately high relative to other racial categories.

We proceed to analyze the findings for the mixed-effects Poisson regression in table 4 to understand what factors predict the number of black homicide victims by police.

The analysis provides evidence that indicates the percentage of black population, the black-white dissimilarity index, the spatial proximity index, percent poverty, and the percentage of black officers are all significant predictors of the number of police homicides of black people. The variable former Confederate state is not statistically significant, suggesting that areas containing a high rate of police homicides of black people are not specific to former Confederate states. Although Jim Crow was more prevalent throughout the former Confederate states, our findings reveal no relationship between police homicides of black people and former Confederate states. This supports Goldsmith and Blakely's (2010) checkerboard analogy of segregation in that predominantly black and white spaces are likely to be segregated by

Table 2.

Correlation Matrix

Table 3.

Number of Homicides by Police Relative to Population, 2013–2017

adjacent neighborhoods. The next set of statistically significant findings provide indications of this trend. We find that the percentage of black population (significant at p < .01) and the black-white dissimilarity index (significant at p < .01) are positively associated with the number of police homicides of black people. A one-unit increase in the percentage of black population increases the incident rate of the dependent variable by a factor of 0.041, or 4 percent (p < .01), holding other variables constant. A one-unit increase in the dissimilarity index increases the incident rate of the dependent variable by a factor of 0.024, or 2 percent (p < .01), holding other variables constant. These findings reveal that cities that have higher populations of black people and areas characteristic of greater racial dissimilarity across neighborhoods are associated with a higher number of black people killed by police. However, although the spatial proximity index is significant, it displays a negative relationship with the number of black homicides by police, although significant at the p < .10 level.

Meanwhile, the correlation matrix shows that the violent crime rate (0.11) and the murder rate (0.29) display a somewhat weak positive correlation with the number of black homicides

Table 4.

Mixed-Effects Poisson Regression Results, Change in Number of Black Homicides by Police, 2013–2017

by police (see table 2). The regression results show that these two variables are not statistically significant when controlling for other variables in the model. These nonsignificant findings coupled with the significant variables for the percentage of black population and the black-white dissimilarity index indicate that police homicides of black people are not driven by high crime rates. Rather, these findings suggest that black communities continue to bear a disproportionate brunt of police homicides and that racial stigmata may underlie a reason for such high figures.

The variable for percentage of black officers is negatively associated with the percentage of black people killed by police (significant at p < .05). A one-unit increase in the percentage of black officers decreases the incident rate of the dependent variable by a factor of 0.031, or 3 percent (p < .001), holding other variables constant. This indicates that having more black officers in the ranks of the police department may reduce the number of police homicides of black and African American residents. These findings lend support to some research on representative bureaucracy (see, for example, Nicholson-Crotty, Nicholson-Crotty, and Fernandez 2017) that suggest having more black officers leads to a decrease in police violence or mistreatment of people of color. In other words, adding more black officers to the personnel ranks could serve as a potential route for creating "safer" spaces in predominantly black and African American communities. However, we interpret these estimates with caution because the data for the percentage of police officers were available only for 2013 and do not account for changes in personnel from 2014 to 2017.

Meanwhile, we found that the poverty rate and the spatial proximity index display a negative relationship with the number of deaths of black citizens. Although the correlation matrix indicates that there is a weak but positive association between poverty and black homicides by police (0.16), the regression displays a significant negative relationship when controlling for other variables. Also, spatial proximity displays a somewhat weak positive relationship, whereas the regression results suggest a significant negative relationship. These findings may be partly explained using Devon Carbado's (2017) police violence model. In this model, Carbado argues one of the dynamics of predatory policing is social control. Other studies exploring poverty in the context of policing have discussed how individuals in targeted social groups may have less access to the protections of law enforcement as a crime victim; that is, these communities may experience underpolicing (Black 1976; Gross and Mauro 1989; Sampson, Raudenbush, and Earls 1997; Pare, Felson, and Ouimet 2007). As Carbado presents, black people, as members of a vulnerable racial group, experience social control policing through practices of spatial management, monitoring, and the directing of their movements (2017). The apparent mobility of black individuals with greater economic means may, for police, challenge preconceived notions of belongingness, thus leading to greater incidents of police violence. Hence the negative relationship may be another indicator of social control policing toward black people.

DISCUSSION

The results of our statistical analysis offer insight for policy implications indicating that race and residential segregation are significant indicators of police killings of black people in the United States. These findings align with Brad Smith (2003), who determined that in cities with one hundred thousand or more residents, racial threat (measured by the proportion of black residents) was significantly related to police-caused homicides. Additionally, these results reinforce the thesis of Cornel West (1994) and the theories of Anderson (2011, 2012, 2015), respectively. Findings support the notion that race still matters and that the iconic ghetto represented by black bodies, perceptions of crime, disorder, and dystopia associated with black spaces, and the negative connotations of blackness continue to endure as powerful sources for stereotype, prejudice, discrimination, and state-sanctioned violence. Black peo ple are more likely to be killed by police in areas where dissimilarity is greater, for example, in more segregated communities. For black people living in white spaces, spaces that are not implicitly associated with crime, their proximity to whiteness may offer increased security that those who live in black spaces do not have. Conversely, these individuals may also have to contend with their neighbors alerting police of their whereabouts because they may be deemed suspicious and out of place. In these regards, black people in the United States continue to navigate social control, as experienced by violence and intimidation by police officers and agencies, similar to their ancestors who navigated lynching mobs and white supremacist organizations.

Black people and black spaces are vilified and victimized, and the marker of "blackness" is instinctively conflated with race and criminal proclivity. Yet, much as the insightful title of Ralph Ellison's (1952) seminal work Invisible Man suggests, the disparate pain and suffering of black people in black spaces seem invisible—out of sight and out of mind of the body politic. To paraphrase Ellison, black people and black spaces represent figments of the public's imagination: Both are invisible because the public and the public servants charged with protecting and serving refuse to see them. Checker-board segregated communities and the purposeful isolation of these communities from more affluent ones allows this willful ignorance to go unchallenged. Even in the twenty-first century and its prevalence of camera phones and social media, as well as the influx of videos capturing the murder of black men and women by police, the refusal to see persists. What has emerged are what Geoff Ward (2016) has poignantly theorized as microclimates of racial meaning—areas plagued by racial violence—both in the American past and present. This is due to the lasting legacy of race-based stigma and the consistency of racial violence through informal or formal mechanisms of social control pre- and post-Plessy.

Although we find black spaces are significantly associated with the killing of black people in the United States, these spaces are not specific to former Confederate states. Segregation and Jim Crow were an explicit part of the Confederacy, though not exclusive to it, yet the relationship between police killings of black people and former Confederate states is not statistically significant. The nonsignificance of these findings goes against common perceptions about the concentration of discrimination in southern states and illustrates the pervasiveness of these practices across the country. These findings challenge discourse that aligns oppressive policing squarely with the South and realigns the narrative in a broader context relating the omnipresence of state violence against black people as a national phenomenon.

Social narratives and political arguments for tough-on-crime policies regularly link high rates of violent crime with a need for increased police presence in communities. Our findings reveal an interesting lack of a statistically significant relationship between police killings of black people and violent crime rates. The policing of black communities and police homicides, as evidenced by our research, are disproportionately higher for black people in the United States. This finding reinforces the pervasiveness of race as an indicator for police killings rather than for the rates at which violent crimes are committed. The nonsignificant finding may also offer insights for policing agencies and local governments on the current allocation of policing resources in black communities.

So, as we consider the future, where do we go from here? Roy Austin and Hiroko Dodge propose that "it is the reality blacks perceive that influences their discontent [with the police]. They [blacks] may base their expectations on the past and present of racial inequality and discrimination" (1992, 594). This notion offers important policy and practical implications for local law enforcement, especially considering the contemporary, polarized context of U.S. society and the need to formulate and implement inclusive policies and practices. In essence, local law enforcement agencies need to acknowledge their past as functional, and in some respects, intentional actors and instruments of state-sanctioned violence to support the spatial face of racism before and after the Plessy decision. Law enforcement also needs to acknowledge the impact these actions and inactions have had on present-day perceptions and realities. Such an acknowledgment needs to be accompanied with the recognition that the present state of discontent with police is not solely because of a few rogue officers, but more so reflective of the organizational culture of U.S. policing institutions as a whole. Most important, these institutions need to take advantage of the opportunity to leverage and learn from the past to create a brighter future that is just and constructive.

Meanwhile, one finding that lends guidance for improvement is in the negative relationship between the percentage of black officers and the number of black people killed by police. This finding may suggest that the presence of black officers in the department and in the community has some role in reducing state-sanctioned violence. Hiring more black officers may integrate nontraditional perspectives into policing policies and practices. Further, a critical mass of black police officers and their collective voice may function as advocates in advancing the interests of black communities within policing organizations while discouraging colleagues from engaging in social control policing that results in higher numbers of police killings. The broader implications are that local law enforcement agencies need to be more creative and aggressive in their attempts to diversify their departments and to repair the damage done by centuries of past police practices that supported a system of race- or ethnicity-based segregation and oppression. Likewise, police agencies and individual officers need to be sensitive to the perceptions that persist.

Franklin Zimring (2020) argues that hundreds of killings each year of civilians by police reflect a power and expertise vacuum and is a problem of governance. He highlights how procedural and jurisdictional problems coalesce and limit the effectiveness of governmental efforts to save the lives of civilians subject to lethal police force. We join Zimring in advocating for more oversight and authority by state governments and their federal counterpart over local police forces, as well as creating the infrastructure for more research that will result in action and administrative reforms to address this issue. These efforts to fill in the governance gaps may have a positive impact on policy, practice, and the profession, and more importantly, save the lives of people.

Moreover, police agencies need to be intentional to effect change in a positive, constructive, and coproductive way. They need to become more aware of the lived experiences of those in which they are tasked to serve and be more responsive to the ever-evolving landscape of U.S. society and the resulting expectations that accompany it. These changes call for embracing a more community-oriented and culturally competent approach to policing. One approach is to promote the cocreation and coproduction of public safety by partnering with residents as they become involved as active participants in all stages of the recruitment, selection, training, promotion, and sanctioning process for police officers. Similarly, police departments need to become more responsive and transparent by embracing civilian review boards. Such an aggressive, planned, and purposeful approach would affect and adapt professional practices and services to better align with the changing landscape and resulting expectations of U.S. society.

Further, racial justice advocates suggest a shift in public budgeting policies that "defund the police." This movement asks not to eradicate policing agencies, but instead for more equitable budgeting policies by reallocating public resources from police departments to other local government and community agencies to support community safety and wellness. The defund movement is, at a fundamental level, aligned with procedural justice and fairness, which "requires administrators to ensure the due process rights of all individuals and to treat all equally" (Johnson and Svara 2011, 167). Further, procedural justice speaks to the idea of fairness that resolves disputes, allocation and reallocation of resources (possibly additional ones), the desire to actualize positive organizational change, and the facilitation of better relationships between the public and police. To assist in these efforts, agencies, organizations, and activists can leverage materials from the President's Task Force on 21st Century Policing (2015), guidelines offered by Community Oriented Policing Services, and the New Era of Public Safety Toolkit offered by the Leadership Conference on Civil and Human Rights (2019), as their guides.

CONCLUSION

Neil Websdale poignantly notes "remarkable and disturbing historical homologies among the control of African slaves in the American colonies, the regulation of freedmen and freed-women during nineteenth-century Reconstruction and Redemption, and the close surveillance and punitiveness directed at blacks [and other minorities] in postindustrial inner cities" (2001, 8). He highlights the strained and interactive relationship that continues to exist between the African American and black communities and local law enforcement agencies. Tall, well-rooted, and established trees do cast long shadows. This reality presents a challenge and an opportunity for the profession of policing because these issues are long standing and deeply rooted. Local law enforcement today needs to understand and acknowledge how the historical legacy and practices of the police profession as instruments of state-sanctioned violence and social control still affect the perception and participation of black communities in collaborative problem-solving and the coproduction of public safety and order. Creative approaches that engender confidence and stimulate the support of historically marginalized and Othered communities can work to begin to reconcile the damage that has been done to foster a symbiotic relationship between black communities and local law enforcement. Out of the historical and contemporary darkness, light can emerge.