How not to Report the Data on Police Shootings

When it comes to policing in the United States, emotion more than reason drives the discussion. The result is Americans continue to suffer as good solutions increasingly become difficult to implement.

To understand complex issues like crime and policing, evidence based approaches and data are important.

However, unbeknownst to many, data can also be used to obscure issues, omit common sense facts, or overplay the significance of certain scenarios or outcomes. In this post I discuss how the media, researchers, and pundits rely on data to obscure crime and policing in the United States.

For example, after looking at the data, one author writes:

In the year 2021 alone, the police have killed a total of 996 people in the United States — while this is a high number of fatalities, it’s even more shocking to learn that 27% of those who were killed by police were Black people, even though they only make up 13% of the United States population (Police Violence Map)

Note: This article is not available online, but from her citation, she must have drawn the data from the Police Violence Map.

The data presented is poignant and shocking as the author says. However,I will highlight two ways in which the author has either tried to obscure a relevant issue or over emphasized something that may not be particularly important.

The author aims to show that black people in the United States suffer more from police violence in any given year than white people, or individuals from other races.

However, while black people were shot by police officers at a rate 27%, which is more than a quarter of the people murdered in that year, it still leaves 73% of non-black people murdered in that year. It takes simple arithmetic to show that most of the people shot by the police in a given year were not black. I know some will argue that Asians, natives, and Hispanics constitute a significant portion of the 73% and they too suffer from systemic racism as people of color. But still, the author draws us specifically towards the plight of black people. These other ethnicities also never make news and the whole world is always fixated on black people. So I will only focus on them here.

As shown, the author seems to get wind of the fact that many more non-black people are killed in a given year than black people. She then goes ahead to make the issue about black murders by the police more serious by adding black people only comprise 13% of the population. That new addition is what I believe bears the ultimate point she is trying to make.

Generally, she implies if the police weren’t biased, black people would be shot by the police at a rate equal to their frequency in the population. She also believes people commit crimes at a rate similar to their ethnic frequency in the population. Both are not true.

Let’s consider a facetious example:

Imagine we have twenty children in a room and twenty sweets in a bag. According to the author, if these children were to steal the sweets, then each would steal a sweet, and not two. But we know that’s false. Out of the twenty children some won’t steal a sweet while others would steal as many as five sweets. Imagine further that out of the twenty children, there are six with green hats, ten with blue hats, and four with red hats. Out of the ten with blue hats, only four of them steal sweets, while two of the six with green hats steal sweets. Lastly, three out of the four with red hats steal sweets. From this hypothetical situation, it is already clear the rates of theft among the three groups is different. It is also readily apparent that the children with red hats have a higher rate of theft (75%) than those with blue hats (40%) even though they are fewer in the population.

That example helps us understand crime and even police shootings in the United States. If a parent were to spank children who steal sweets, those in red hats would receive spanks at a higher rate (75%) than their frequency in the population (20%). Similarly, blue hats would receive spanks at a rate of 40% even though they comprise 50% of the population. If these were police shootings, the rate would be as it appears for spankings.

To wit, just because black people comprise 13% of the American population does not mean the police murder rate in that community will be 13%. Similarly, just because white people comprise 67% of the American population does not mean their rate of murder by the police will be similar. Therefore, even though the author presents accurate facts, she does not consider an obvious question, “what if black people commit more violent crimes which makes it more likely they will encounter the police?”

Since the author is hell-bent on showing that the police shoot more black people than whites:

1. She uses data to obscure the fact that more non-black white people are killed by the police in any given year than blacks. This is not a rate, just pure numbers (996–269=727 non-black). Sometimes converting raw numbers into percentages can obscure common sense. Also think about this, if the police used race to decide who to shoot, how did they end up killing more non-black individuals?
2. She also overemphasizes the idea that police shootings are supposed to be proportionate to population frequencies, that is, despite some children stealing more sweets, she thinks all of them need to be spanked equally. This is the argument that most advocates use especially when you point out the obvious fact #1. The fact that black people are killed at a higher rate says more of their over-representation in crime despite being underrepresented in the population.

Black homicide rate was 39% despite comprising 13% of the population (FBI, 2020)

I believe her intention was to show that the police in America are racist towards black people. However, if it requires a police officer to calculate whether to shoot or not to shoot a suspect based on the frequency of the offender’s race in the population, then there’s a high likelihood the police officer would make the wrong judgement risking both his life and that of innocent victims.

This is not to mean that police officers are free to shoot innocent civilians, or that police brutality against black people is acceptable. What I mean is that black people are being killed by the police at a rate equal to (or less than) their criminality. The same is true of arrests and incarceration. For racial differences in crime rates, check the FBI data set readily available online.

Innocent or guilty

A good way to think about the relationship between police shootings and the murder of black people is to begin by asking ourselves how many of the 269 black people murdered by the police in 2021 were innocent. Many people believe the police target and kill black people for no other reason other than racism. George Floyd’s death is a more recent example of what is believed to be murder by racist police officers. While, Floyd’s death was in 2020, I am sure you’ve had nothing of the 269 black people murdered in 2021. Why didn’t their deaths make any news?

Of the 269 black people killed in 2021 by the police, high chances are 99.9% of them were criminals who posed a direct threat to the police. The same is true of the remaining 73% of white, Latino, Native American, and Asians killed by the police in the same year. Suffice to note that, most of the murders that make news are of innocently murdered black people. But these cases are a handful. From this Wikipedia list, only two black people are listed for the 2021 entry. Also worth noting is that, these few deaths are not enough to make the conclusion that police officers are racist. In most cases, poor judgement on the police’s side could lead to a fatal shooting of an innocent civilian. I found this reading particularly insightful when it comes to explaining police shootings.

To sum, the point of this essay was to show how crime data can be presented in a way that obscures common sense facts and observations. There are many ways in which data and statistics can be used to fool people. In this post, I only focused on a small piece of data that tried to explain police shootings but ended up drawing wrong conclusions. The author not only failed to mention crime rates of various ethnicities but also overemphasized the importance of racial frequencies in explaining police shootings.