Place not Race: Understanding Police Homicides in the USA
While BLM activists and their supporters have claimed that police across the country systematically kill Black civilians, there is scant evidence supporting these assertions. Except for anecdotal accounts gathered from news reports, there is little data to support claims of widespread killings of civilians — of any race — at the hands of the police. In fact, based on data gathered by the Washington Post, most civilians killed were not black. However, an analysis of the location of civilian fatalities at the hands of police reveal that civilian deaths by police tend to cluster in a relatively limited number of communities across the country. Rather than a widespread, systematic problem, civilian fatalities at the hands of law enforcement may reflect variations in state and local law enforcement policy, practices, and operating context.
Due to media coverage, many Progressive liberals (both Black and White) believe that Blacks are the primary victims of police homicides. The tendency for the media to nationalize these uniquely tragic incidents has led many activists to believe that police abuse is ubiquitous and systematic. For liberals, police homicides reflect the country’s past and present institutional racism. However, by almost any metric, police homicides are a rare, unique occurrence — more like “Black Swan” events. New research shows that place, not race, would be a better explanation for the number and kind of police homicides.]
The death of a citizen at the hands of a police officer while performing official police duties is called a police homicide. The term homicide is not a normative but descriptive term — i.e., it describes an event where one individual killed another individual. Homicides can be justified or unjustified; intentional or accidental. Police homicides simply chronicle the number of civilians killed by police officers — for any reason — while performing official duties. Progressives believe that police homicides are ubiquitous and that blacks are the primary victims. A recent study by Skeptic Magazine found that liberal and very liberal individuals believe that 56% to 60% of victims of police homicides are black. However, from 2015 through 2019, only 24% of the victims of police homicides were black.
While the data do not support liberal and very liberal citizens’ estimates of Black victimization by police, some Progressives argue that given their population share Blacks are disproportionately victims of police homicides. This suggests that police should determine whether to use deadly force based on population percentages — i.e., if Blacks are roughly 13% of the US population, then they should only be 13% of the police homicide victims. An alternative argument is that each police-citizen interaction is unique and is driven by the context and circumstances surrounding the encounter. For instance, according to the FBI’s Uniform Crime Report (UCR) in 2019, Blacks were 27% of all arrests (about twice their population share). Moreover, Blacks were 51% of all individuals arrested for murder and nonnegligent manslaughter. In short, Blacks have offense rates for all and serious crimes that are much larger than their share of the population. Using arrest rates there is a strong argument that Blacks are underrepresented amongst police homicide victims.
In the Skeptic Magazine survey mentioned above, liberal, and very liberal citizens not only incorrectly assessed the overall share of Black police homicide victims, but substantially overestimated the number of unarmed blacks killed by police annually. Liberals and very liberal citizens estimated that police kill 1,000 or more unarmed blacks annually. According to the Washington Post and the Mapping Police Violence Project, in 2019 the number of unarmed blacks killed by the police was between 13 and 27 — a huge order of magnitude different from political Progressives’ estimates. As with other assessments of the prevalence of police homicides, these distorted estimates seem to be driven primarily by media accounts which treat every incident of a police officer’s use of fatal force against a black citizen as a national story.
Examining police homicides as a local phenomenon, shows that police-civilian contacts that result in a civilian death are both rare and disproportionately concentrated in a few locations. Based on research conducted by this author of the Washington Post database on fatal police-civilian encounters that controlled for county population size, from 2015 through 2020, 78% of large US counties (2240 counties) had 2 or fewer police homicides. During this period, 22% of these same large US counties accounted for over 80% of all civilian fatalities by police officers (5,814 civilian deaths). The level of geographic concentration is even more dramatic when counties are classified by whether they had 11 or more, 3 to 10, or 0 to 2 police homicides. Using this method to classify police homicides, roughly 4% of large counties in the US accounted for 49% of all police homicides. High levels of geographic concentration are also evident when controlling for race: 41% of all police homicides with Black victims occurred in just 45 (2.0%) large US counties; and 22% of police homicides with White victims occurred in just 45 (2.0%) large US counties. Moreover, there was considerable overlap between counties where Blacks and Whites were police homicide victims.
Figure 1. Location of Counties with 5 or more Police Homicides by Race 2015–2019
Figure 1 illustrates the degree of geographic concentration of police homicides and the racial makeup of fatalities in each county. Counties with 5 or more police homicides appear to be disproportionately concentrated in the West, Southwest, and parts of Florida. Police homicides appear to be rare in the Northeast, New England, Mid-West, and the upper plains region. The geographic concentration of police homicides provides little support for Progressive activists’ and commentators’ that argue that police violence against citizens, especially Blacks, is widespread and endemic. On the other hand, Center-Right analysts and commentators may also be incorrect in concluding, that despite the media narrative, police homicides are extremely rare events and as such not really an issue. The available data indicate that police homicides are clearly an issue in some places around the country. For example, from 2015 through 2020, Los Angeles County had over 240 police homicides, while Maricopa County had over 160, and Harris County (Houston City) had more than 80. This contrasts with St. Louis County, Mo with only 8 police homicides: Boulder County, Co with 6 or Bronx County, NY with 5.
The unfortunate focus on trying to explain police homicides by using the race of the police homicide victim and the race of the police officer involved in the incident as causal factors has limited exploration of other explanations. The data indicate that place rather than race may be a more salient explanation for the prevalence of police homicides. As such, the major policy question should be: what is happening in those counties and regions with relatively high numbers of police homicides compared to those counties with no or few police homicides and what can we do about it?
