Joseph Cesario's Lab
Overview of Current Research on Officer-Involved Shootings
This page describes the recent research in our lab on police use of deadly force, and specifically the possibility that police officers in the U.S. are more likely to fatally shoot Black vs. White citizens.
We have tried to provide some perspective on this important topic in two different ways: (1) by analyzing real-world shooting data, and (2) by collecting data in controlled, laboratory experiments. Each line of research is described below.
I. Analyses of Real-World Shooting Data
Until recently, it has been difficult (if not impossible) to accurately study police shootings at the national level due to the poor databases maintained by the federal government. In the last few years, The Guardian and The Washington Post have filled this void by building more complete databases of fatal police shootings. We have also compiled our own unique, large database of officer characteristics, collecting information about the race, sex, and years of experience of officers involved in fatal police shootings.
We have analyzed police shootings from these databases to answer two questions: (1) Are police more likely to fatally shoot Black citizens vs. White citizens? and (2) Does officer race, sex, or experience predict being more likely to shoot Black citizens vs. White citizens?
(1) Are police more likely to fatally shoot Black citizens vs. White citizens?
Click here for a copy of this paper and click here for a detailed description of this research.
The short summary of this research is the following:
-
The standard way of testing whether there are racial disparities in fatal police shootings is to ask whether the proportion of Blacks shot by police is greater than the proportion of Blacks in the U.S. population. (For example, ~30% of those shot are Black, but Blacks make up only ~15% of the population, thus there is a racial disparity.) This answers the question, "Are Blacks or Whites more likely to be shot given population proportions?"
- We argue that it is misleading to compare whether a group is shot more or less than their U.S. population proportion.
- Instead, we argue that a better benchmark is to compare the proportion of those shot to the proportion of those who commit crime, as a proxy for exposure to the police, as the situations in which police use deadly force are overwhelming crime-related. This answers the question, "Are Blacks or Whites more likely to be shot given each group's involvement in crime?"
-
We find no evidence, at the national level, that officers show racial bias against Blacks in the decision to use deadly force.
(2) Does officer race, sex, or experience relate to the race of the citizen shot?
We have collected perhaps the most complete database of officer characteristics to date, compiling officer data on all fatal police shootings in 2015.
Click here for a copy of this paper and click here for a description of this research.
The short summary of this research is the following:
-
We found no evidence that the race of a police officer related to the race of a citizen shot. In other words, having White officers at a shooting did not make it more likely that a Black citizen was shot, compared to having Black officers at a shooting.
- Instead, race-specific violent crime rates strongly related to the race of the citizen shot. The greater number of crimes committed by Whites in a county, the more likely the person shot was White. The greater number of crimes committed by Blacks in a county, the more likely the person shot was Black citizen. And the greater number of crimes committed by Hispanics in a county, the more likely the person shot was Hispanic.
- Consistent with our other work, this work suggests that the single most effective solution to reducing fatal police shootings is to reduce crime.
II. Our Experimental Research
To complement our analyses of the real-world shooting data, we also conduct controlled laboratory experiments to test for racial bias in officers' deadly force decisions. We collect data by having officers go through our immersive shooting simulator, during which they watch first-person videos and make decisions to shoot firing a modified handgun.
We have collected data from almost 1,000 officers so far. These include sworn officers as well as several samples of police recruits. The main manuscript for these data is currently under review.
In addition, we have a paper published (Johnson, Cesario, & Pleskac, JPSP) on the role of dispatch information in the decision to use deadly force, as assessed by laboratory "First Person Shooter Tasks." Results show that dispatch information eliminates any racial bias in this task.