American Sociological Association

ASA Footnotes

A publication of the American Sociological AssociationASA News & Events
July/August 2020

The Demography of Police-Involved Homicides

Michael Esposito, Postdoctoral Fellow, Institute for Social Research, University of Michigan
Frank Edwards, Assistant Professor of Criminal Justice, Rutgers University
Hedwig Lee, Professor of Sociology, Washington University in St. Louis

Michael Esposito

The recent murders of Breonna Taylor and George Floyd by law enforcement have ignited one of the largest social movements in U.S. history. Citizens across the country have taken to the streets to demonstrate against excessive, yet seemingly routine, police use of force—and, more broadly, against racism that is deeply embedded in our social, economic, and cultural systems.

With widespread demonstrations introducing the notion of racial injustice in policing to new audiences, it is perhaps worthwhile to recapitulate the demographics of all too common fatal police-civilian interactions. As W.E.B. Du Bois—the founding father of social demography—demonstrated through his seminal research, simply describing inequalities via careful demographic methods can be a powerful tool for informing effective social change. In this spirit, we thought that it might be helpful, for the current moment, to share demographic estimates from our work that outline the scope of fatal police-violence in the United States.

Risk of Being Killed by Law Enforcement by Gender, Race, and Age

In a study published late last year, we used data provided by Fatal Encounters to estimate how one’s risk of being killed by U.S. law enforcement varies by age, gender, and race/ethnicity.

Frank Edwards

Our analysis finds that about 52 per 100,000 males and 3 of every 100,000 females will be killed by police. For comparison, the lifetime risk of an individual—male or female—being killed in a firearm-related homicide is about 350 per 100,000 population (which includes police-involved shootings). We find that police are responsible for approximately 8 percent of all homicides with adult male victims every year.

Our analysis also demonstrates sizable racial disparities in the risk of being killed by police. According to our estimates, approximately 1 of every 1,000 Black men/boys can be expected to be killed by law enforcement. (Note that 1 in 1,000 is also the mortality rate of measles which, according to Rhea Boyd, pediatrician and Director of Equity and Justice for The California Children’s Trust,  is “... a risk deemed so deadly, it has near constant public health surveillance and prevention.”) For American Indian and Alaska Native men/boys, lifetime risk of being killed by police is about 1 in 2,000 and for White men and boys, this risk is approximately 1 in 2,500. Despite rates among women being nearly 20 times lower than rates among men, racial disparities in chances of being killed by police are still quite pronounced among females: we estimate that Black and American Indian/Alaskan Native women and girls are about 1.5 times more likely to be killed by law enforcement than their White female counterparts.  

We also find that risk of being killed by law enforcement fluctuates across the life course. Young men and women, particularly between the ages of 20 and 29, have the highest rates of police-involved killings among all age groups. For young men, police violence ranks as the sixth leading cause of death, after accidents, suicides, other homicides, heart disease, and cancer. Risk

Hedy Lee

of being killed by law enforcement is particularly pronounced among young men of color, with police being responsible for 1.6% of all deaths to Black men between the ages of 20 and 24.

Geographic Variation in Risk

Our analyses show that rates of fatal civilian-police interactions in the United States are not equally distributed across the United States. In an analysis of adult males, we find that where an individual resides is tied to their chances of being killed by law enforcement (2018).

Perhaps most interestingly, we find that race-specific rates of being killed by police vary among clusters of U.S. states. Indeed, among adult Black males, overall risk of being killed by law enforcement is highest in Pacific (e.g., Washington, California) and West North Central (e.g., Minnesota, Missouri) states—where more than 3 per 100,000 men are estimated to be killed by law enforcement each year. Among Latino adult males, risk is by far the greatest in Mountain states (e.g., Arizona, New Mexico), where police are estimated to kill around 1.9 per 100,000 Latino adults each year. White adult male risk appears relatively more stable across regions, but is somewhat elevated in Mountain states and individual states like Oklahoma.

Racial disparities in risk are also quite heterogeneous across the U.S. landscape. Black-White disparities appear most exaggerated in Middle Atlantic (e.g., New York, New Jersey); East North Central (e.g., Michigan, Indiana); and West North Central states—where adult Black males are estimated to be up to 8 times more likely to be killed by law enforcement than their White male peers. Disparities in risk between adult Latino and White males remain fairly consistent across the entire country.

While less pronounced than state-based variation, we also find differences in risk across metropolitan types. Individuals living in large, central metropolitan areas are generally at the highest risk of being killed by police, at about 1 death due to police per 100,000 population each year. However, we find that about two-thirds of all police-involved killings occur in suburbs (e.g., places like Ferguson, MO), smaller cities and rural counties. Indeed, we find that in rural areas, law enforcement officers are responsible for more than 10 percent of all homicides with adult male victims.

Moving Forward

According to Fatal Encounters data, U.S. law enforcement kills more than 1,000 people per year or more than 3 people each day. When these figures are situated alongside rates from advanced industrialized democracies, it becomes clear that the U.S. is unique in the violence enacted by its law enforcement institutions. In other countries, such as England or Germany, police kill fewer individuals over the course of a few years than U.S. police do over the course of a few days. In addition, our work highlights how racial/ethnic disparities in risk are quite severe—and how police are a principal source of early mortality among young males, especially young African American men. Altogether, the current demography of police-involved homicides suggests that the U.S. has an unignorable issue in how it performs policing.

We and other sociologists (e.g., Sewell and Jefferson 2016; Sewell, Jefferson, and Lee 2016; and Vitale 2017) suggest that police-involved deaths, and other forms of violent interactions with law enforcement, are not solely the result of the inadequacies in criminal justice policy. Instead, the United States’ high rate of police-involved homicides is rooted in our health, education, and economic structures. Ensuring that communities are safe from excessive, fatal police use of force will thus require upstream interventions, including educational and economic investments; improved housing infrastructure; robust social services and improved access to health care. Directly addressing policing is only the tip of the iceberg.

More personally, following civically engaged scholars who study inequality (Hunter 2018; Itzigsohn and Brown 2020; Morris 2017), we believe that it is critically important to share our research with the public and highlight the inequities that plague our society. We cannot look away, but must act. In the words of the late U.S. congressman John Lewis,
When you see something that is not right, not fair, not just, you have to speak up. You have to say something; you have to do something.

We hope that we are doing our part.