Following the lines of oppression: a data analysis of police shootings in the US

A Data Journalism project by Anna Uras

By the end of May 2020, the brutal homicide of George Floyd by Minneapolis police officer Derek Chauvin fueled historic protests in the US and all over the world. The Black Lives Matter movement took the streets of Minneapolis and rapidly spread through the whole Country, shouting the last worlds of George Floyd: “I can’t breathe!”.

After months of Covid-19 dominating the news, newspapers and Tv channels were suddenly flooded by updates on the protests, accompanied by endless commentary and debates.

In their coverage of the fire at Minneapolice police station, BBC reported that

Minneapolis Mayor Jacob Frey called the unrest “unacceptable”, but recognised there was “a lot of pain and anger”.

The most diffused opinion among the (mostly white) commentators was that the anger of the protestors was justified, but they should have expressed it differently.

But what is the right way to express anger? And who decides it?

When we talk about anger, we refeer to the anger of people who are systematically oppressed by the society they live in. Those who looking at the video of George Floyd killing were not only shaken by the brutality of a murder, but by the deep knoweldge it could have been them.

As a feminist, I often find myself trying to explain my anger to men. Mostly, they agree women should not be discriminated. And yet, they say I am too angry, or that I should express my anger differently. That not all men are the same. It is true. Not all men are violent. But all women live the same systematic oppression. Every women reads the news about a women being raped or killed, knowing it might happen to her.

It is difficult to explain something like that. Because ultimately, while we live in the same reality, we do not experience it the same way. As a white women, I can try to explain my anger at male violence, and I should listen to the voices of people who suffer racist violence.

As Tamika Mallory powerfully stated in her viral speech:

Don’t talk to us about the lootings. You are the looters. America has looted black people, America looted the Native Americans when they first came here. Looting is what you do, we learned it form you. We learned violence from you. So if you want us to do better, then damn it, you do better.

The Data on police shootings: an intersectional approach

Data are not enough to understand how different subjectivities experience reality. However, data are not destined to remain cold facts either. Through data analysis it is possible to reduce the impact of individual experience and get a wider perspective. In this case, a look into the data on police shootings allows us to follow the threads of oppression and police violence.

Police shooting in the US (2015 - Now) source: The Washington Post

The above map shows the fatal police shootings occurred in the US in the last 5 years. It does not include all the killings. For example, George Floyd would not results among the victims, as he was not shot.

Race distribution among victims of letal police shootings

However, the picture that emerge from the data shows a reality that is far from that of a world were All Lives Matter, as some claimed in response to the BLM movement. While white people account for 73% of the US population, they merely account for 50% of the victims. On the contrary, 26% of the victims were black (despite only accounting for 12.7% of US population).

Another relevant data is that almost all the people who were shot to death by a police officer were men. While this difference is partially explained by the rates of criminality by sex (man constituted almost 74% of the people arrested in 2012), the difference remains significant.

Relative frequency of Male and Female among victims of police lethal shootings (2015–2020)

Moreover, while roughly 50% of the victims of police shooting among men are white, this figure reaches 63% among women. At the same time, black and hispanic victims respesent respectively 27% and 19% of the victims among men, while they only account for 20% and 12% of the victims among women.

Race distribution of victims among women
Race distribution of victims among men

In short, being black seems to increase the risk of being shot to death by the police. Moreover, being a man strenghten the correlation between race and police brutality. Of course, the relationship between socially rooted discrimination such as those based on gender and race cannot be reduced to a single case. In many other instances, being a women might increase the strenght of racial oppression (for examples, in cases of discrimination on the workplace or sexual violence). However, this seems to be the case when it comes to police brutality.

But what were they doing?

Finally, a question often posed when discussing data on victims of police brutality is What were this people doing to be shot? Could this explain the data?

The answer is no. Not only it does not explain the racial distribution of victims, but it increases the impact of race on the data.

For example, while only 6% of the white people who were shot by the police were unarmed, almost 10% of black victims were unarmed. When it comes to the threat level that the victims of letal shooting represented to the police (as measured by the Washington Post, click here for more on the criteria), roughly 66% of both white and black victims represented a high threat. However, the level of threat was lower for hispanic, asian and native americans.

In conclusion, both the data and the voices of BLM protesters portray the same reality. A reality where being black means a higher risk to be killed by the police, regardless of what you do.

As it’s always true when it comes to violence, the killings merely represent the tip of the iceberg of a systematic oppression, and there is no easy solution. But there will be no solution at all if we don’t stop debating the legitimacy of the anger of black people and start listening to them.

Journalist and student at the UvA University of Asterdam