These shocking cases have focused national attention on what seems like a rash of questionable police killings involving minority citizens. Until now there has been no way of telling if the perception of police bias is due to media bias, statistical flukes, or real patterns of differential policing in minority neighborhoods. A big question hanging in the air:
Is the use of deadly force disproportionately used in Black or Latino communities?
I will answer that question below using newly gathered data. The question can't be answered by standard government statistics. There is no mandatory or systematic reporting of police killings in America. With 17,000 law enforcement agencies acting mostly on their own, fatal encounters are seldom noticed beyond local news accounts, the bulk of these relying on local police statements. There is no public notification requirement when people are killed by police.
That's why it is a big deal that several citizens have begun scanning the internet for local news accounts of police action fatalities and compiling a listing of them on their KilledByPolice.com Website. Shocking preliminary findings showed that in just 23 months of record keeping 2,181 people were killed in police encounters. For some perspective, there were only 70 police killings in the past 90 years in Great Britain.
This analysis only looks at three demographic categories, White (non-Hispanic), Latino, and Black. All other demographic categories are excluded because they are still too few in number for a meaningful analysis. The population totals and fatality data used here are only for these three study groups. This is not to diminish the importance of other demographic groups, especially in the fatality data, but that analysis will have to wait until more data is available. Also, local news accounts often do not report the race, ethnicity, or even the identity of the person killed. This forces us to assume unreported demographic information, if known, would reflect the population as a whole rather than the fatality patterns observed below. This data includes all cases where a civilian is killed in a police action whether accidental, justified, or unjustified.
Below are four comparative bar graphs for each of the US Census regions of the United States. It compares the distribution of the total White, Latino, and Black population in the region with the distribution of people killed in police encounters within the region.
If police action fatalities were random events then the demographic makeup of people killed should reflect the population. The four regional graphs above almost speak for themselves. In every region of the country, Blacks comprise a significantly larger portion of those killed in police encounters than expected based on regional demographics. You will notice that Latino populations are also over-represented among police action fatalities in the West. I will discuss results in each region below and show some of the state data that may be significant in understanding these findings.
In the past 23 months, there was a disproportionate number of police-involved killings in the African-American community in the Northeast Region. There were 197 instances where people were killed in their encounters with police. This is the lowest number in any region, but this is also the region with the smallest population (for these three racial/ethnic categories). In 150 of these cases, the ethnicity or race of the person killed was reported in the press. Blacks make up 13% of the population in the Northeast and 37% of police action fatalities. This is an unrepresentative finding. The numbers for Latinos killed by police actions were slightly under-represented. They represent 13% of the study population and 11% of those killed by police. Because the Northeast has fewer police action fatalities several states have fatality numbers too low for meaningful analysis. The three exceptions are New York, New Jersey, and Pennsylvania.
In Pennsylvania and New York, there were higher numbers of fatalities than in New Jersey and evidence that these were also disproportionately Black lives,
Pennsylvania had 52 police action fatalities, 14 where race/ethnicity was not reported, and 38 others. Blacks make up 11% of the three racial-ethnic groups in this study and 34% of those killed in police actions. Latinos are 6% of this population and 8% of the fatalities.
New York had 56 incidents, 14 where the race wasn't reported and 42 others. Blacks comprise 17% of the study population and 43% of those killed in police actions. Latinos represent 19% of the study population and 9% of the casualties in police actions in the past 23 months.
The number of fatalities in Massachusetts was also disproportionately black, especially relative to the population, but the total number of police action fatalities was too small to be sure of the significance. Massachusetts had 26 police action fatalities of which race was not reported in 7 and 19 others where it was reported.
In the remaining states of Connecticut, Maine, New Hampshire, Rhode Island, and Vermont the total number of people killed by police was 32 with 6 cases where the race was not identified. Of the other 26 cases, 18 were White, 4 Latino, and 4 Black.
In the past 23 months, there were 339 instances where people were killed in their encounters with police. In 255 of these cases, the ethnicity or race of the person killed was reported. Blacks and Latinos each make up 13% of the study population but Black citizens made up 40% of all those killed in the Midwest during police encounters. Latinos were less likely to be killed in police actions. They were just 5% of the casualties. The states that had the most fatalities and largest disparities in the region were Ohio, Illinois, and Missouri.
Ohio's study population here is 13% Black yet 55% of those killed in police actions were Black. No Latinos were reported killed. There were a total of 61 cases where citizens were killed by police actions. The race was not reported in 10 of those cases.
Illinois had 63 citizens killed in police actions. No race or ethnicity was reported in 37% of those cases. Blacks are 15% of the study population but 55% of the police action fatalities. Latinos fared better, being 17% of the population and 12% of the casualties.
In Missouri, there were 49 citizens killed in police actions of which race was reported in 37 incidents. Blacks make up 12% of the study population and 49% of the police action fatalities.
In Wisconsin, Michigan, and Indiana the rate of Black fatalities in police encounters was also higher than expected, but the number of individual cases was too few for individual analysis. The numbers for these three states were combined. The result yielded 87 cases, 61 of which where the race was reported. Of those 61 cases 36 casualties were White, 2 Latino, and 23 Black. Blacks make up 11% of the population in this combined group and 38% of the fatalities in police encounters.
In the remaining states of Iowa, Kansas, Minnesota, Nebraska, North Dakota, and South Dakota the combined total of people killed in police actions was 46 with race not reported in 15 cases. No minorities were killed in North Dakota or South Dakota. Blacks and Latinos were 15% and 11% of the total casualties and Whites were 74% of the total.
The Southern Region is the most populated of the four regions. There are many states with over 30 incidents in 23 months, partly because of higher populations. There were 890 cases where people were killed in their encounters with police. In 652 of these cases, the ethnicity or race of the person killed was reported. The Southern Region has the highest population, the highest number of police action fatalities, and the highest percentage of African-Americans in the population. Blacks make up 20% of the population and Latinos another 17%. White, non-Hispanics are 63% of the total population. Black casualties of police actions made up 40% of the total cases in the prior 23 months. Latinos had 13% of the police action fatalities.
Alabama had 31 police action fatalities in 23 months where the race was identified. Blacks represent 27% of the study population and 55% of the casualties. There were 8 additional cases where the race of the victim was not identified. No Latino fatalities were identified.
In Florida, there were 140 police action fatalities with a very significant 35 cases in which race or ethnicity was not identified. Blacks make up 17% of the study population and 38% of the casualties while Latinos are 23% of the population but only 13% of the casualties.
Georgia had 42 cases with 18 additional cases were the identity of the casualties were not provided. Blacks make up 32% of the state and 55% of the police action fatalities. Latinos make up 9% of the study population and just 2% of the casualties.
Louisiana had 27 police action fatalities where race the race of the casualties were known and 13 where this information wasn't initially reported. Blacks are 33% of the population and 595 of the casualties. Latinos were evenly represented at 4% of both the study population and the casualties.
Maryland had 31 cases in 23 three months plus an additional 12 where the race was not identified. Blacks make up 23% of the population but 68% of the police action fatalities. The highest ratio in the nation. Latinos represent 9% of the study population and 6% of the casualties.
North Carolina had 48 cases and 11 more where the racial or ethnic identity was not reported. Blacks make up 22% of the state and 55% of the police action fatalities. Latinos are 9% of the state and 4% of the casualties.
Oklahoma, with the smallest population in the South, had 45 police action fatalities where the race was identified and another 12 where it wasn't. Just 9% of the population is Black and they represented 38% of the fatalities. Latinos make up 10% of the population and 18% of the police action fatalities.
Tennessee had 31 incidents and 9 more cases where the racial identity was undisclosed. The state is 17% Black and Blacks made up 29% of the fatalities. Latinos are 5% of the population but 0% of the fatalities.
Texas has a large population and there were 153 police action fatalities where ethnicity or race of the casualties was identified. There were 51 other cases where this information was not reported in the local press. That is the highest number in this category. Blacks are 12% of the population and 25% of the police action fatalities. Latinos are 40% of the study population but only 37% of the casualties. The high number of cases where race or ethnicity was reported could substantially change these percentages.
West Virginia is note worth for having no minority police action fatalities. There were only 13 cases with 5 White casualties identified. The state is mostly White.
Turning attention to the West, there were 697 people killed in police encounters in the past 23 months. Of those, the race of those killed by police was reported in 491 cases. In this Region, Blacks make up 6% of the study population and 14% of the police action fatalities. This is the lowest percentage of Black casualties in absolute terms but still over twice what it should be by chance alone. The Western region, on the other hand, has the highest percentage of Latinos in the four regions. It is also the only region where Latinos have a higher rate of police action fatalities than chance would allow. Latinos comprise 33% of the population and 38% of people killed in fatal police encounters.
Arizona had 94 police action fatalities in the past 23 months and the race of the person killed was released in 67 cases. Blacks comprise 5% of the population under study here and 9% of the casualties. Latino fatalities were more disproportional. They make up 32% of the population and 42% of the people killed in police encounters.
California had the most civilian fatalities in police encounters in the country and is also the most populated state. It had 351 police action fatalities of which the race of the casualties was reported in 251 cases. Blacks represent 7% of the population and 20% of the police action fatalities. Latinos are 45% of the population and 47% of the casualties.
Colorado had 43 police action fatalities of which 16 cases provided no information on the race. Blacks are 4% of the population and 7% of the casualties in a police encounter. This is among the most representative of Black demographics for any state among high population states. On the other hand, Latinos comprise 22% of the population and 45% of the police action fatalities. This is the most disproportional rate for this group nationwide.
Nevada had 32 people killed in police encounters of which race was not reported in 12 (38%) of the cases. Blacks comprise 9% of the population and 15% of the casualties in fatal police encounters, while Latinos represent 30% of the study population here but 15% of the fatalities.
New Mexico had only 28 fatality incidents and just 3 cases where the race of the casualties was unreported. Blacks comprise 2% of the study population and 8% of those killed in police encounters. Latinos are 52% of the population and 48% of the civilian casualties.
Oregon had 29 fatal police encounters, 19 in which the race of the people killed was reported. Blacks are 2% of the study population and 5% of the casualties. Latinos make up 13% of the population and 11% of those killed in police encounters.
Utah had 25 police action fatalities, 23 in which race was identified. Blacks account for about 1% of the study population and 9% of the fatalities in police encounters. Latinos are 14% of the study population and 13% of the total, which is essentially a representative proportion.
In Washington State, there were 58 cases where civilians were killed in police encounters. In 23 cases (43%) the race of the persons killed was not reported. Blacks comprise 4% of the study population and 11% of the fatalities. Latinos represent 13% of the population and 17% of the fatalities.
Wyoming had only 4 police action fatalities and all of the casualties were White.
If my hypothesis is to disprove that African Americans are more likely to be the casualties in fatal encounters with law enforcement (the null hypothesis), then this journalistic analysis failed to prove it.
It might still be true that there is nothing to the perception that Black Americans are more likely to die in police encounters, but the evidence here strongly suggests the contrary, they are a disproportionate class of people killed in police encounters in virtually every region of the country. It can be said that this isn't true in every state, but that is of little comfort to minority communities.
This review of this newly collected data supports the suspicions that minority communities may be subject to differential police tactics or over-policing. The depth and breadth of these finds are disturbing. Blacks are simply more likely to be killed in police encounters in every area of the country and a majority of the state. Latinos are also more likely to be among the fatalities in the West. Fatalities in these police encounters are also overwhelmingly men. The gender basis of these findings have yet to be studied or even noticed, One might be tempted to think this higher rate of Black men being killed in police encounters is connected to so many young Black men being unemployed and on the streets, yet the average age of those killed by police is 36.9 years. This is almost exactly the average age in the U.S. population. This means there is no statistical bias due age which rules out a young black male theory.
Another factor some folks bring up is the high crime statistics in minority neighborhoods and a greater chance that minority folks act like criminals. First of all, there are very few crimes that carry a death sentence, which is was is administered by law enforcement where there are safer ways to apprehend a suspect. And crime statistics are mostly sensitive measures of policing policies, tactics, training, law enforcement priorities, and the culture or attitudes of those doing the policing. If you target any given group to enforce laws their crime statistics will climb. A secondary result of economic disparity and heavy-handed policing is the contempt for the law that is often seen in poor and minority neighborhoods. The latter follows as a consequence of the former in nearly all cases.
As for minority folks acting like criminals, who is making that judgment, and what is the evidence?
I have reviewed dozens of specific recent cases at this point and can say that some of these police action fatalities are completely justified but in most cases, there aren't enough detail to form any judgments. There are also many cases where it seems law enforcement had other options to avoid the use of deadly force. Very few cases involve independent investigations of the incident or judicial reviews. They rely on mostly local police accounts with no eyewitnesses or video cameras. The police are policing themselves.
To the best of my knowledge, this is the first look at what is probably the most comprehensive collection of information about police action fatalities. This is more a journalist exercise than a scholarly study and this information badly needs a more scholarly treatment. Still, I believe the original question asked here can be tentatively answered in the affirmative:
The use of deadly force is disproportionately used in the Black community in every region of the country and in the Latino community in the West.