OnPolicing Blog

Origins of officer-involved shootings: Analysis of data reported to police via 911 calls reveal opportunities to reduce violent outcomes

November 25, 2020

NPI icon

National Policing Institute

NPI icon

National Policing Institute

Through the cooperation of more than 50 of the largest law enforcement agencies across the U.S. and Canada, our research team was granted access to the most detailed dataset ever collected on fatal and non-fatal shootings by officers while on duty. Covering the period from 2015 through 2018, these data cover over 1,000 fatal and nonfatal shooting incidents, including characteristics regarding the reason for and circumstance of the encounter, the location, and the officers and the other individuals involved. While differences in data availability and policies impacted our ability to collect every detail across all agencies and incidents, the data provide important insights on these statistically rare, but massively important, events. In particular, the data shed light on an often overlooked aspect of the events that likely contributes significantly to the problem and yet remains understudied and unaddressed.

Although traditional and social media content may lead some to believe that officer-involved shootings most often begin with encounters involving a traffic or vehicle stop, our data reveal that in the large agencies we worked with and those providing sufficient data to assess the origins of these incidents, over half of these events began with a community member calling 911. Slightly less than half of the encounters began by an officer-initiated activity (e.g. initiating a traffic stop).

Focusing on community member calls to 911 that resulted in officer-involved shooting incidents, we found that 69% of callers reported that the individual was or may be armed with a firearm. However, investigations later revealed that in 35% of incidents where callers reported a firearm, there was no firearm recovered or person armed with a firearm. Although the data reported by callers was accurate in 65% of shooting events, a 35% error rate is unacceptable in matters of life and death. We should point out that conversely, in cases where the caller did not indicate the presence of a weapon (about 31% of citizen-initiated events), the data show that 36% of individuals involved in these events were later found to have been armed with a weapon of some kind, according to the data collected.

Call takers and dispatchers have no reliable way to know if a caller is honest or accurate in what they report, but they do have a professional duty to share all relevant information with responding officers and to do so quickly. For some types of calls, such as child abductions, active shooters, missing persons, etc., protocols exist on what call takers should ask or determine. There is no protocol for what to ask when a caller reports an armed person and thus it is largely up to the natural ability of the call taker to ascertain important details and to assess the quality of the information being provided, assuming the operating environment, staffing, call volume, etc., allow for these abilities to be leveraged fully.

So what leads to this inaccurate reporting from callers in particular that may trigger a series of events and decisions that could result in a life being taken unnecessarily and also jeopardizing the safety of bystanders and responding officers? Recent incidents may provide important insights:

On April 24, 2020, the Austin Police Department received a 911 call that a man and woman were using illegal drugs and that the man was armed with a gun, potentially holding the woman against her will. Eight officers responded to the call and were clearly prepared to confront an armed person they had encountered previously. The person was ultimately shot and killed by officers as he tried to flee the scene after initially cooperating. The person told officers he was not armed, but the 911 call indicated otherwise. After the shooting, officers learned that the person was not armed and the woman with him was not being held against her will. Media reports indicate that the 911 call and caller may be under investigation for making a false report.


On May 21, 2020, the Phoenix Police Department received a 911 noise complaint followed by another call indicating that the noise complaint (screaming) had escalated into a physical fight. Two officers responded to the apartment provided by the caller. When the person answered the door with a firearm in his hands, both the officers and the person appeared to be surprised based on body camera footage released by Phoenix Police. As the person reacted to the presence of the police, the police reacted to the presence of the gun and the movements of the person, killing him on the scene. The person’s girlfriend, who was in the apartment with him, later reported that they were playing video games which may have gotten loud. In the 911 call recording, also released by Phoenix Police, the caller acknowledges “It could be physical. I could say yeah if that makes anybody hurry on up. Get anybody here faster.”


In these incidents, it is fairly obvious that the callers’ reports were not accurate. But in many other cases, such as the 31% of shootings that turn out to not involve an armed person, we know they were wrong, but we don’t know why. Is it that they want the police to respond faster, they want the police to take a more aggressive approach, an honest mistake, or could it be implicit or explicit bias playing a role? Or a combination? The May 25th Central Park incident between a dog walker and bird watcher is an example of how bias and racism may play a role in calls to 911 as the police are weaponized through false reporting. In that case, a disagreement between the two resulted in the white individual threatening and then following through to make a 911 call to report threatening behavior and attempted assault by an “African American man.” Both reports were later determined to be false, resulting in her prosecution.

What we do know is that the information provided by callers—whether truthful or not—can change how officers respond and could “prime” their decision making in ways that make negative outcomes more likely. Research has found that “when the dispatched information was erroneous, it contributed to a significant increase in shooting errors.” (Taylor 2019) Additionally, research has found that the call-taker process of risk appraisal may escalate caller uncertainty which may lead to increased priming and negative outcomes. (Gillooly, 2020).

As we consider policing reforms, we should not eliminate from consideration the potential problems that we, the public, may contribute to knowingly or otherwise. We must learn more about the frequency of these events, who is involved and impacted in terms of demographics and motivations, best practices for mitigating, and explore ways that technology may be able to assist call takers and dispatchers in assessing data quality and reliability. We can save lives if we do this and for this reason among others, understanding and reducing the use of deadly force will remain a top priority of the National Policing Institute.



Gillooly, JW.  How 911 callers and call‐takers impact police encounters with the public: The case of the Henry Louis Gates Jr. arrest. Criminol Public Policy.  2020; 19: 787– 803. https://doi.org/10.1111/1745-9133.12508

Taylor PL. Dispatch Priming and the Police Decision to Use Deadly Force. Police Quarterly. 2020;23(3):311-332. doi:10.1177/1098611119896653

Leave a Comment

This site is protected by reCAPTCHA and the Google Privacy Policy and Terms of Service apply.

Disclaimer: The points of view or opinions expressed in this article are those of the author(s) and do not necessarily represent the official position of the National Policing Institute.

Written by

NPI icon

National Policing Institute

Strategic Priority Area(s)

For general inquiries, or to submit an essay for consideration, please contact us at info@policefoundation.org.

Strategic Priority Area(s)

For general inquiries, please contact us at info@policefoundation.org

If you are interested in submitting an essay for inclusion in our OnPolicing blog, please contact Erica Richardson.