Is It True Unarmed Blacks Are More Likely to Be Killed in Police Encounters?

13th Amendment 2nd Amendment African American Lives American Social Casteism Constitution cultural sensitivity implicit bias Police Reform Public health Policy Race and health Racially biased policing Uncategorized

Over the years the United States has been a country known for racially motivated violence. This issue has moved into the forefront of the political discourse in the past few years. In 2020 George Floyd’s death with videos of Derek Chauvin’s knee on Floyd’s neck on social media sparked protests nation-wide. Many believe these protests along with the poor handling of the SARS-CoV-2 pandemic led to Trump’s loss in 2020. The issue of police violence against minorities especially blacks is however not new. Many ask “Is It true unarmed Blacks are more likely to be killed in police encounters?”

Death or violence at the hands of police is one thing that many blacks worry about a lot. This problem got some national attention after the Rodney King video in the early 90s. With the availability of smartphone videos the visibility of these violent acts have gone through the roof. In the past two decades as we inaugurated our first African American President we also saw more reports of racially mediated violence against blacks by the police.

What is The Washington Post Fatal Police Shootings Database?

In 2015 The Washington Post began capturing all fatal shootings by on duty police officers in the US. This means this data would not include deaths such as Freddy Gray, Sandra Bland, George Floyd and Eric Garner. These deaths and many more which did not involve a firearm were not included in this data. Amber Guyger’s off-duty shooting of Botham Jean is not included in this database.

From January 1st 2015 to January 16 2021, 5,960 killings have been logged in this database. The post collected data from news reports, social media posts and updated this using police records. Many believe this data has more entries than the Federal Bureau of Investigations (FBIs) data on the same issue. Among those with data on race 50.8% were non-Hispanic White, 28.3% African American, 18.6% Hispanic, 1.8% Asian and 1.5% Native American. In this post I would compare the the relative proportions of African Americans in the general populations of these regions and states with their representation in shootings.

For the purpose of this report, unarmed means unarmed or without any specific weapon reported based on the police report. I used simple statistical tests to determine if the proportions of blacks versus non-blacks among the unarmed killings were due to chance. In other words I assumed police officers are just as likely to kill an unarmed black person as a non-black.

Comparing Rates of Fatal Shooting to State Racial Proportion

In the first set of analysis I compared the black proportion of state populations to their representation in fatal shootings. The US map below shows the data by state with states having death ratios of 1 – 6. This is the proportional ratio of blacks involved in shootings to the general state population. Some of the high numbers may not be of any relevance since most of the numbers at the state level were very small. Grayed out states did not have enough shootings over the period for any meaningful analysis.

Relative Proportions of Blacks in Fatal Police Encounters by State
The states with the highest death ratios were 6 for UT and 4 for both IL and WV. Some states that appear gray may have higher death ratios but did have total fatalities less than 50 over the 7 year period and were excluded from analysis. Data source: WAPO database

States with a death ratio of 2 and above have more than a twofold ratio of fatal encounters compared to the black population proportion. This may suggest a bias or just plain aggressive over policing of black neighborhoods in these states. Most may just be an artefact of inadequate data and lack any relevance. Of the 50 states only 18 have death ratios of one or lower. In the remaining 32 states and the District of Columbia a black person has at least twice as much risk of dying in a police encounter as is to be expected if there was no excess shooting of blacks by the police. This assumes all relevant parameters are similar across racial groups.

Police Arrested Dylan Roof Alive But Jacob Blake Was Killed, Was This Due To Chance or Is Something Sinister Going On?

Everyone knows a story of a black shooting that make most people ask why on looking at the facts. In the same vein a white guy who kills many people is arrested unharmed. Whenever I hear such stories I ask myself why oh why? So when I downloaded this data I asked myself what simple statistical test will answer this question. So I used the statistical test that we call chi-squared tests of proportions. This test looks at the proportions of different groups by another parameter. So I calculated the odds of a black person being involved in an unarmed police shooting. I calculated this at the national, regional and then finally at the level of states. I will only report key findings.

The results of this analysis is heartbreaking if you are black. Nationally a black person is 1.7 times more likely to be involved in an unarmed fatal police encounter compared to all other races. The p-value for this test was < 0.001. This means that there is a less than 1 in a 1000 likelihood that this finding was a chance finding. At the level of US regions the odds are represented on the graph below.

Odds of an Individual Involved in an Unarmed Police Shooting Being Black By Region of US
In the Southeast and Southwest an unarmed police shooting is 2.8 and 2.7 times as likely to involve a Black individual as opposed to another race. The differences in race was only significant for these 2 regions with a p-value < 0.001 Data source: WAPO database

The highest odds are in the Southeastern and Southwestern Regions of the country. In these regions a report of an unarmed police shooting is 2.7 – 2.8 times more likely to involved an African American. This difference was statistically significant suggesting that it is highly unlikely that this is a chance finding. The odds of this being a chance finding is 1 in 1000. The other regions had lower odds; 1.6, 0.7 and 1.7 for Northeast, Midwest and the West respectively; these were not statistically significant.

Only 10 US states had enough numbers of unarmed shootings across all races to make any state level analysis meaningful. These states are; AZ 19, CA 66, CO 11, FL 30, GA 11, LA 12, OH 16, OK 15, PA 10, TX 28, WA 11. Among these states only Arizona, Florida and Texas showed a significant tendency for African Americans to be the victims of unarmed police shootings. The data is shown in the table below.

State Level Data on The Odds of Unarmed Black Police Fatalities
StateOdds of unarmed black fatalityNumber of unarmed Fatal shootingsBlack % of unarmed fatalitiesAll fatalities armed and unarmed P-value for odds
Arizona 4.819 21.0%2780.01
Florida’s high percentage of African American involvement in unarmed police shootings is especially concerning for a state with only 16.9% Black Population
What Are The Causes of This Problem

The data collected by The Washington Post shows that in the United States from January 2015 till now we still have a problem with excessive force by police against African Americans. There are many theories why this practice continues. Using the information that we have we can only make educated guesses on the root of this problem. One theory is that this is a remnant of the Jim Crow era black codes. There may be some truth in that. The regions with the highest odds of an unarmed black person being killed by the police in this analysis were the Southwest and the Southeast. These were mostly prior slave states.

Others believe that most White Police officers who shoot unarmed African Americans are actually acting out of fear. Such officers are responding in the midst of a very heightened physiological response to a perceived threat. This theory presupposes that exposure to language and imagery suggesting black violence conditions such individuals. Such a preconditioned person is more likely to act violently when faced by a black person irrespective of the real threat posed.

What Are Some Appropriate Solutions?

The interpretation on this analysis supports both theories discussed in the prior paragraph. The data suggests a high likelihood of pervasive systemic anti-black attitudes and beliefs driving these killings. To help curb these tragic deaths among both blacks and whites, officers will need better desensitization training related to fear and xenophobia. Cultural sensitivity and immersion experiences will be useful for specific units that have more contact with minorities. Police departments must appropriately address racially motivated behaviors of all kinds and not only focus on fatalities and egregious acts.

Our nation needs healing and this healing can best be achieved in a society that is willing and ready to talk about race.

By Dr. Leonard Sowah an Internal Medicine Physician in Baltimore, Maryland

Feature Photo: Black Lives Matter Demonstration in Berlin, September 2020. Creator: Montecruz Foto Copyright: Montecuz Foto


A physician providing primary medical care to patients across the lifespan