On September 14, the 13-year-old Tyre King was shot and killed by a police officer, Bryan Mason, in Columbus. According to the police, Tyre was shot after running from officers who were looking into an armed robbery and pulling out a BB gun that looked similar to a real weapon.
Tyre was black and Bryan white, fitting into a broader pattern of police killing of black males. Despite calls for calm from city officials, there is outrage over this shooting, with protesters calling for homicide charges against the police officer involved. Police union leaders, in turn, are defending the officer, saying he acted appropriately.
However, we have very limited information about what occurred. Definitive judgments on either side of whether the officer was justified or not should make us question the objectivity of those making such statements. Most likely, those folks are falling into confirmation bias, a typical thinking error that our minds tend to make when we let our strong personal beliefs override the evidence available.
So how can we think effectively about whether the shooting was justified? Rather than a binary yes or no approach, we can apply the principles of probabilistic thinking to gain the most accurate understanding of the situation possible based on the evidence available. Probabilistic thinking is a research-based strategy that involves placing percentage estimates on what we believe and updating our beliefs based on the evidence as it emerges.
First, let’s consider Columbus police policy: officers should use deadly force only when they have a reasonable belief that they are defending themselves or others from serious injury or death. Since Tyre was not threatening anyone else at the time, we can safely assume that Bryan was defending himself when he shot Tyre. Thus our belief of whether the shooting was justified rests on whether we believe the officer perceived himself to be in immediate deadly danger.
The officer claims that after he chased Tyre into an alley, the boy pulled out his gun. Since Bryan believed it was real, he shot Tyre. If Tyre indeed did pull out his gun and hold it in a threatening manner, I would personally put a very high probability of the shooting being justified.
However, the only one who was in the alley and came out alive was the officer. In many previous cases, especially involved white officers shooting African-Americans, the accounts of the police officers involved in questionable shootings have proved false.
Imagine what would happen to Bryan if Tyre was actually not threatening him. What if Tyre was running away when he was shot? What if Tyre did not pull out his gun, but simply had it on him? What if Tyre was trying to throw the gun away to not be found with it? Not only would Bryan’s career be ruined, but he would be tried and likely convicted for killing Tyre. He has a very strong incentive to lie. This lowers my percentage probability of Bryan’s account being accurate.
So does other evidence. For instance, Tyre is described by those who know him as a well-behaved kid who was into sports. He also attended Bible study. Given Tyre’s background and that Tyre knew it was a BB gun and that it would not do him any good at all to shoot it at the officer, I place a low probability on him actually threatening Bryan with his gun. Of course, we can’t discount the possibility that Tyre panicked and did do what the officer said he did, but it seems unlikely.
Some additional evidence is an independent investigation conducted on behalf of Tyre’s family by Francisco J. Diaz, MD, FCAP, a practicing medical examiner in the County of Wayne, Michigan, and a Consultant Forensic Pathologist. According to Dr. Diaz : “it is more likely than not that Tyre King was in the process of running away from the shooter.” This further undermines the probability that Bryan’s account is fully genuine.
More broadly, this shooting has to be understood in context of tense relations between the police and African-Americans over racial profiling. A report from the Federal Government’s Bureau of Justice Statistics indicates that black people were three times as likely to be searched as whites in a traffic stop. Certain police officials wave away data suggesting unfair racial profiling by claiming that blacks are more likely to be involved in criminal activity than whites. Yet a study just released by the Center for Policing Equity corrects for racial difference in criminal activities. The study’s outcomes show that African-Americans are more than three times as likely to suffer from police use of force compared to whites, for everything from mild restraint to gunshots.
Now, this does not mean that white police officers explicitly discriminate against black people. Research shows that all of us suffer from some degree of implicit bias, deeply ingrained negative attitudes associated with certain groups or markers of social identity. The large majority of white Americans – including police officers – are implicitly biased against African-Americans. These studies should further decrease our probabilistic estimate that Bryan was justified in shooting Tyre.
There are some research-based methods of debiasing which are likely to reduce implicit bias. One great technique is de-anchoring, where instead of going with our intuitive reaction, which is highly likely to be biased, we adjust our behavior based on research. For instance, if a police officer knows that black people are over three times as likely to experience police coercive force, she should think thrice before using force on someone who is black. Although further research is needed on how to deliver such training effectively, if Bryan had gone through this sort of training, I would have higher confidence in Bryan’s account.
We can’t discount the possibility that Tyre panicked and did do what the officer said he did. However, based on the current information available, I would put a 20% estimate on the shooting being justified. Of course, your own estimate of the weight of each piece of evidence may vary, and I am very open to updating my beliefs based on additional evidence.
How much more healthy and productive would our discourse be if we placed probabilistic estimates on our beliefs and disputed how much weight each piece of evidence should have? We can then have a much more nuanced and accurate approach to such polarizing and emotional questions.