Police Use of Deadly Force


Randall G. Shelden and Pavel Vasiliev[1]


Historically, there had been a low level of public concern about the rate of deaths that occur in police/citizen interactions. Similarly, there was almost no research on lethal violence by the police. As Zimring notes:

The local nature of governmental responsibility for police has a pervasive influence on what types of information are available about police killings and where it is available. The agency of government that collects information about police killings is the local police department. And any quality control on information about police killings must also happen at the local government level. There may also be a perception in government, media, and public opinion that each violent episode involving police use of deadly force is an individual drama rather than a part of a larger expression of governmental policy. (Zimring, 2017:  11–12).

When a white police officer shot and killed unarmed Michael Brown on August 9, 2014 in Ferguson, Missouri, the ensuing riots prompted national media coverage of shootings by police officers (and the riots on April 19, 2015 after Freddie Gray’s funeral intensified coverage). A month earlier, Eric Garner had died from a chokehold. Through July 6, 2016, there was extensive coverage of the deaths of minorities during police/citizen encounters: Levar Jones, Laquan McDonald, Akai Gurley, Tamir Rice, Walter Scott, Freddie Gray, Samuel Dubose, Alton Sterling, and Philando Castile. The issue of police use of deadly force became a subject of academic attention and public debate (Schenwar, Macare, & Price, 2016). There was a transformation from viewing the killings as singular events to representative incidents in an institutional pattern. In addition, the focus changed from the regulation of police conduct to a question of civil rights (Zimring, 2017).

There were no answers to questions such as: How many persons are shot and/or killed by the police each year in the United States? Geoffrey Alpert, a criminologist at the University of South Carolina, said it was a national embarrassment that there were no records of how many times officers fire their weapons (Kappeler & Potter, 2018). The absence of comprehensive data on the number of arrest-related fatalities was conspicuous. Calls for the development of a national database on police use of force have been made repeatedly (Alpert, 2016; Kane, 2007; Koper, 2016).

The only data available was collected for other purposes, and much of the information relied on voluntary submissions by police agencies. The Centers for Disease Control, the Federal Bureau of Investigation (FBI), and the Bureau of Justice Statistics (BJS) collected some information. The numbers from these sources differed widely in any given year and from state to state.[2] For example, the Arrest-Related Deaths (ARD) program designed by BJS captured from 36 to 49% of deaths that occurred during the arrest process from 2003–2009; in 2011, the percentage increased to an estimated range of 59–69% (Banks, Couzens, Blanton, & Cribb, 2015). The Supplementary Homicide Reports (SHRs) compiled by the FBI as part of the Uniform Crime Reports (which do not include federal law enforcement agencies) captured 46%.The President’s Task Force on Twenty-first Century Policing recommended increased collection of data on officer-involved shootings. In October 2016, the Federal Bureau of Investigation announced the launch of a centralized national data collection including nonlethal uses of force. The National Use of Force Data Collection was slated to begin in 2017 (Barnett-Ryan, 2016).

The majority of police shooting victims are white (52%) but African Americans (32%) are shot disproportionate to their numbers in the population (DeGue, Fowler, & Calkins, 2016). The fatality rate is 2.8 times higher for blacks than whites. Researchers analyzed the 991 fatal shootings identified by the Washington Post in 2015 (Nix, Campbell, Byers, & Alpert, 2016). After adjusting for the age of the person shot, mental illness, and for the crime rate in the neighborhood where the shooting occurred, the researchers found unarmed black men were shot at twice the rate as unarmed white men. The study found that the police exhibit implicit bias by falsely perceiving blacks to be a greater threat than non-blacks to their safety. Violent crime levels do not increase the likelihood of a lethal shooting. Detroit and Newark have high levels of crime but lower rates of police shootings (Mapping Police Violence, 2016). The probability of an unarmed African American being shot by the police is about 3.49 times the probability of being white, unarmed, and shot by police (Ross, 2015). Crime rates or racial differences in offending do not explain the heightened odds, as “racial bias in police shootings is not reliably associated with crime rates and not related to the difference in race-specific crime rates” (p. 10). Racial discrepancies in fatalities were more frequent in larger metropolitan counties with low median incomes, higher income inequality, and a higher percentage of African American residents. Police shootings are less frequent in areas with the highest levels of criminal violence than in those with middle levels of violence (Klinger, Rosenfeld, Isom, & Deckard, 2016).

In 2015, 25% of the deadly force incidents occurred in 59 of the 60 largest city police departments (there were no fatalities in Riverside, CA). The highest rate of police killings were in Bakersfield, Oklahoma City, Oakland, Indianapolis, Long Beach, New Orleans, and St. Louis. African Americans were 41% of the victims. Fatalities in 41 of the large police departments were disproportionately of African Americans. In 14 departments, African Americans were the only fatalities; in 5 departments, whites were the only fatalities (Mapping Police Violence, 2016).

In 1985, the Supreme Court ruled in Tennessee v. Garner that a fleeing felon must pose a significant threat of death or serious harm to innocent bystanders or to officers for a shooting to be legally justified. In 1989, the Court ruled in Graham v. Connor that deadly force is justified if the officer reasonably believed he or others were in imminent danger. In 2015, the Supreme Court ruled in Mullenix v. Luna that officers are immune from lawsuits unless a shooting was unequivocally unreasonable. Police officers have considerable discretion in deciding whether to fire their weapons; the officer’s perspective is generally accepted. When officers kill civilians, they report that they were acting in self-defense (Kappeler & Potter, 2018). There is a tendency to err on the side of using deadly force, reinforced by the fact that officers are rarely punished for their behavior. The militarization of the police has encouraged the use of force. SWAT teams are increasingly used for executing warrants, credit card fraud, and other low-level offenses. The display of such force escalates tensions and teaches officers to be wary of interactions.

Amnesty International (2015) found that state laws, where they exist, do not comply with international standards on the use of lethal force by law enforcement officers. International law enforcement standards require that force of any kind be used only when there are no other means available that are likely to achieve the legitimate objective. If force is unavoidable, it must be proportionate to the objective. Nine states and the District of Columbia have no laws on the use of lethal force by officers. No state requires that lethal force be used only as a last resort. Only eight states require that a warning be given before lethal force is used. Only 3 states require that officers not subject bystanders to risk when using lethal force. Only 2 states have statutes regarding training in the use of lethal force. Accountability is severely lacking. The officer’s police agency usually conducts the investigation before a local prosecutor, who relies on good working relationships with the police, decides whether to file charges.

In an analysis of thousands of fatal police shootings from 2005 to 2014, Philip Stinson, a criminologist at Bowling Green University, and The Washington Post found that only 54 officers faced charges for fatal police shootings (Kindy & Kelly, 2015). Charges were filed only in cases where victims were shot in the back, the incidents were captured on video recordings, other officers testified, or there were allegations of a cover-up. If a case goes to trial, jurors tend to see officers as those who enforce laws rather than break them (Kappeler & Potter, 2018). Winning a conviction is difficult. Locke Bowman (2016), clinical professor of law at Northwestern University School of Law, explains:

Almost without exception, the police officers have escaped punishment—the system has had their backs. . . . Our laws have been written intentionally to make it hard to hold police accountable. A lawyer friend of mine explained it to our mutual client: “The system makes rules to protect the people in the system.” (p. 21)




Alpert, G. P. (2016). Toward a national database of officer-involved shootings: A long and winding road. Criminology and Public Policy, 15(1), 237–242. 

Amnesty International (2015, June 17). Deadly force: Police use of lethal force in the United States. New York: Author. 

Banks, D.,Couzens, L., Blanton, C. & Cribb, D. (2015, March). Arrest related deaths program assessment: Technical report, NCJ 248543. Washington, DC: Bureau of Justice Statistics. 

Barber, C., Azrael, D., Cohen,A., Miller, M., Thymes, D., Wang, D. E., & Hemenway, D. (2016). Homicides by police: Comparing counts from the national violent death reporting system, vital statistics, and supplementary homicide reports. American Journal of Public Health, 106(5), 922–927. 

Barnett-Ryan, C. (2016). Development of a national data system to collect officer-involved shootings and use of force resulting in serious bodily injury or death. Paper presented at the 2016 meetings of the American Society of Criminology in New Orleans, LA. Bartlett, D. L. & Steele, J. B. (1992). America: What went wrong? Kansas City: Andrews and McMeel. 

Bowman, L. (2016, July 18). Delays in McDonald case foster despair in Chicago. Chicago Tribune, p. 15.  

DeGue, S., Fowler, K. A., & Calkins, C. (2016). Deaths due to use of lethal force by law enforcement: Findings from the national violent death reporting system, 17 U.S. states, 2009–2012. American Journal of Preventative Medicine, 51(5), 173–187. 

Kane, R. J. (2007). Collect and release data on coercive police actions. Criminology and Public Policy, 6(4), 773–780. 

Koper, C. S. (2016). Advancing research and accountability on police use of deadly force. Criminology and Public Policy, 15(1), 187–191. 

Kappeler, V. & Potter, G. (2018). The mythology of crime and criminal justice (5th ed.). Long Grove, IL: Waveland Press. 

Klinger, D., Rosenfeld, R., Isom, D., & Deckard, M. (2015). Race, crime, and the micro-ecology of deadly force. Criminology and Public Policy, 15(1), 193–222. 

Loftin, C., Wiersema, B., McDowal, D., & Dobrin, A. (2003). Underreporting of justifiable homicides committed by police officers in the United States, 1976–1998. American Journal of Public Health, 93(7),1117–1121. 

Mapping Police Violence Project. (2016). 2015 police violence report. Retrieved from: http://mappingpoliceviolence.org/2015/ 

Nix, J., Campbell, B. A, Byers, E. H., & Alpert, G. P. (2016). A bird’s eye view of civilians killed by police: Further evidence of implicit bias. Criminology and Public Policy, 16(1), 1–32. 

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, 0(00), 1–34. doi: 10.1371/journal.pone.0141854. 

Schenwar, M., Macare, J., & Price, A. Y. (2016). Whom do you serve, whom do you protect? Police Violence and Resistance in the United States. Chicago, IL: Haymarket Books. 

Zimring, F. E. (2017). When police kill. Cambridge: Harvard University Press.


[1] This is taken from a chapter in the third edition of Controlling the Dangerous Classes: A History of Criminal Justice in America (forthcoming).  Long Grove, IL: Waveland Press.


[2] The differences are due to definitional, organizational, methodological, and population coverage issues, and both the National Vital Statistics System and the Supplementary Homicide Report underreport the number of deaths (Loftin, Wiersema, McDowal, & Dobrin, 2003). The National Violent Death Reporting System is more complete (Barber et al., 2016) Later iterations of the Arrest Related Deaths Program captured up to 69% of deaths (Banks, Couzens, Blanton, & Cribb, 2015).Journalistic and open-source projects provide higher estimates.