Content Warning: Discussion of Police Violence.

This project was made by Mari Reid and Sam Lavin, students at St Olaf College, for a final project in their MSCS 150 class. All data used in this project comes from https://mappingpoliceviolence.org/. In an effort to not just reduce tragedies into data points we urge readers to explore https://theirnames.org/ to gain a human perspective into this delicate matter. This website by Harvard lists not only the names of people murdered by police but also provides more information about their individual stories.

Introduction

The murder of George Floyd by Officer Derek Chauvin was a tragedy that shook not only Minneapolis but further exposed the world to the systematic racism buried deep into America’s very identity. Something about this murder in particular finally awoke many people to the racism in their own neighborhoods. With data on police violence we predict to find evidence that a disparity exists in police killings of Black people and we will be able to show that racism by the police is not a problem localized to the American South but is widespread throughout the nation.

Data Description

In this data analysis we will be looking at several different variables to look at how the rate of Black people being killed by police compares to that of white people, as a whole and by geographic regions. The data was collected from 2013-2020.

Variables, Alphabetically:

Armed/Unarmed Status: Sorts victims by if/how they were armed when they were murdered. A victim could be coded as: Allegedly Armed, Unarmed/Did Not Have a Weapon, Unclear or Vehicle (someone killed while they were driving dangerously to the public). Factor.

Black-White Police Violence Rate Disparity: The disparity in rates of murders by police of white and Black people. Calculated by (Average Annual Rate of Black Murders)/(Average Annual Rate of White Murders). Unitless. Numeric.

Number of Black People Killed: A count of all Black people killed by police, by state. Numeric.

Number of People Killed: A count of all people killed by police, by state. Numeric.

Percent of the Population which is Black: The percentage of the population that is African American, by state. Numeric.

Percent Victims who are Black: The percentage of all victims killed that were Black, by state. Numeric.

Race: Sorts victims by their race. Includes, Asian, Black, Hispanic, Native American, Pacific Islander, Unknown Race, White. Factor

Rate of Black Murders: The average rate of Black people murdered by police in a year. Calculated by: (Black People Murdered from 2013-2020)/(Black Population)*(1000000/7 years). In people/year. Numeric.

Region: Sorts states into five geographical regions. Includes, Pacific, Mountain, Midwest, Southeast, Northeast. Factor.

Results

This graphs shows the total number of all people killed by police in a state from 2013-2020. The color of the state reflects the number of murders that police committed. A state with few to no killings would be gray, and as the number of killings increase the state becomes lighter blue, with a bright sky blue representing the maximum number of killings. This map shows that police have killed many people in every state over the last eight years and police violence is a national problem that is present everywhere. This issue, although significant for all Americans is a particularly significant threat for Black Americans.

This graphs shows the total number of all Black people killed by police in a state from 2013-2020. The color of the state reflects the number of Black people murdered by the police. Color is used in the same way as in the previous map; a state with few to no killings would be gray, and as the number of killings increase the state becomes lighter blue. The map reveals the vast number of Black people that have been murdered over the last eight years, and highlights how specifically police killings of Black people is a national problem that needs to be addressed.

## # A tibble: 1 x 4
##   Total_Pop Pct_Pop_Black People_Killed Pct_Killed_Black
##       <dbl>         <dbl>         <dbl>            <dbl>
## 1 322903030          12.3          8770             25.0

This bar graph shows the number of the victims who belong to each race. It shows that more white people have been murdered by police than any other group, and Black Americans have been murdered the second most. This graph is useful for providing a sense of scale to the problem, but on its own does not reveal how outside factors, such as racism, affect the rate at which any one group is being murdered. With all outside factors being removed you would grouped well represented in the whole population to be equally well represented on the graph. The table shows how that Black people are tiwce as present among those killed by police than they are in the whole population.

This scatterplot plots fifty-one points, representing the states and DC, by the percent of the state’s population that is Black and the percent of police violence victims in that state who are Black. There is also a maroon line that represents the equation y=x. (This was created with the geom_abline() function) A point on this line means the percentage of Black people murdered corresponds with the percentage of Black people in that state’s population. In a world without external factors you would expect the points to be clustered around this line. Points above the line represent a state in which the percentage of Black victims is larger than the percentage of Black people in the general population. With so many points above the line it shows that external factors, such as implicit bias and systematic racism, are causing Black people to be murdered in disproportionate amounts.

This bar graph sorts all the victims of police violence from the last eight years by their race, and shows the percent of victims of each race that were armed, unarmed, were using a vehicle as a weapon, or if their status is unclear. The bars are clustered by race, and the color of the individual bars correspond to the armed/unarmed status. The height of the individual bars represent the percent of victims in each race group that match the corresponding armed/unarmed status. This graph shows that not only are Black individuals being murdered in disproportionate amounts but a greater percentage of those murdered are completely unarmed. This further suggests the presence of outside factors, like personal bias, that leads to unarmed Black individuals being perceived as threatening by police officers.

This Cleavland plot shows the differences in the rate at which Black people are murdered by police in different regions of the US. The points on the lines represent the average number of Black individuals murdered by police per year in that region. The line extending from the point indicates the region where specific state’s in that region rates are likely to fall. This plot shows how Black people are being murdered at high rates, although they vary, across the whole country, not just in former confederate states that are considered “more racist”. Without further information, it is possible to think that Black people are murdered at such high rates in a particular region because all people are being murdered at high rates in that region. The next plot will compare the rates at which Black and white people are murdered in these same regions to show their is a disparity.

This series of boxplots highlights the large disparity between the rate that white people are killed by police and the rate that Black people are killed by police. In each box, the end of the bottom “whisker” represents the minimum (that is not an outlier), the bottom of the box represents the 25th percentile (as in 25 percent of points are at or below this line), the line that cuts through the box is the median, or 50th percentile, the top of the box is the 75th percentile, and the end of the top “whisker” represents the maximum, (that is not an outlier). The dots beyond the “whiskers” represent outlier points. The position of the boxplot on the y-axis shows how extreme the Black-White rate disparity is in that region. A position at 5, for example, means that a Black person in that region is five times more likley to be murdered by the police than a white person. The plot shows that there is a large disparity across the US, no matter the region, because a median value as low as two, is still two times larger than it should be. It is significant to note that the regions with the largest disparity are the Northeast and the Midwest, and the smallest disparity exists in the Southeast. This shows that the implicit bias and systematic racism that allows such disparities to exist is not a “Southern Problem” but is one that exists throughout the country, especially in regions that white people have historically have deemed as “not racist”.

Discussion

This data exposes just how large a problem police violence is in America. It has cost thousands of Black, and other, individuals their lives. With Black people being killed at a rate three times higher than white people, and with a greater percentage of those people being killed unarmed, it is clear that there are outside factors causing this disparity, being personal bias and systematic racism. However, it should be noted that the largest limitation of our analysis is that we cannot prove the cause for the disparities we have observed. We used our common sense and previous knowledge to make the reasonable assumption that racism impacted the discrepancies, but the data itself does not account for any direct measure of racism. This problem exists not only in the Southern states but throughout the entire nation. Since the scope of this problem is so large it is something that every individual should be aware of, not just BIPOC communities that are the most affected by this injustice. This isn’t stopping either, even after the activism that stemmed from the murder of George Floyd. What this all tells us is that racism, both systematic and individual, is all too prevalent in America and that changes must be made to end the massive amounts of injustice in the American police system.

Our data analysis is only showing the tip of the iceberg for this matter. We were not able to utilize all of the massive amounts of data collected by https://mappingpoliceviolence.org/. Further exploration would be valuable into the conditions in which the people were killed, what is the impact of body cameras, what police departments have the most racial disparity and whether the officer was convicted. If we were to do this again, we would have liked to examine whether similar trends exist for other minorities, particularly with Hispanic and Native American victims. Although, Black Americans face the highest rates of murder by police overall, other minority groups also are killed at disproportionate rates and it would be fascinating to dive into how prevalent a factor racism is in other cases.

Conclusion

Through our data analysis we were able to show that massive numbers of Americans have been murdered by police in the last eight years, and that a disproportionate number of those victims are Black and unarmed. This is a problem not just localize to the South but exists throughout the country, especially in the Northeast and the Midwest. These findings support our hypothesis that a disparity exists between the rate at which Black and white people are killed by police across the whole country.

There is no easy solution to this problem that plagues America. Racism is at the very core of the police system and for American culture. The best thing an individual can do is to support Black voices advocating for change. This can be done in simple ways, such as, by listening what others have to say, showing support at a protest, and contacting your representatives. Systematic racism will only be rooted out when the system changes. So, it is up to us to call on our lawmakers to dismantle the police institution as we know it and for justice to be served to those who take innocent lives. This won’t change in a day, but identifying the problem is the first, of many, steps towards fixing it.