Skip to main content
ABC News
The Police Departments With The Biggest Racial Disparities In Arrests And Killings

The police response to the insurrection at the U.S. Capitol, in which at least 39 law enforcement officers participated, has renewed conversation about racial bias and links to white supremacist groups among police officers. As state and local reforms continue to be implemented across the country, the new presidential administration has promised to step up the federal government’s role in fighting racism in policing.

[Related: The Police’s Tepid Response To The Capitol Breach Wasn’t An Aberration]

The Biden administration plan to “root out” bias within the nation’s local police departments relies in large part on the U.S. Department of Justice’s authority to launch civil rights investigations and require the adoption of new use of force standards, accountability structures and other systemic changes in places where its investigators find a “pattern or practice” of unlawful or discriminatory policing. While these reforms are not a panacea, cities that have been the subject of Justice Department investigations have seen reductions in police shootings following these interventions.

Which cities should the Department of Justice investigate? According to the department’s civil rights division, it prioritizes investigating places with “patterns of unlawful use of force; unlawful stops, searches and arrests; and racial discrimination.” Data can help point us to places where policing practices look the most problematic by these criteria.

That data is incomplete — the federal government does not collect comprehensive data on police stops, searches or use of force. But it has published data on over 8 million arrests made in 2019. Additionally, nationwide data on killings by police is available from 2013 through 2020, via the Mapping Police Violence database. Police departments with the highest rates of arrests and use of deadly force, especially when applied in racially disparate ways, should prompt the federal government to investigate whether these outcomes are due to discriminatory or unlawful conduct.

To determine which police departments might fit these criteria, I calculated per capita arrest rates and police killings rates for some of the largest U.S. cities’ police departments (cities with more than 500,000 residents, according to the U.S. Census).1 There are methodological differences to consider when interpreting this data. For example, Hispanic people are coded as Hispanic in the Mapping Police Violence database but coded racially as white or Black in the FBI’s Uniform Crime Report. This can underestimate the extent to which there are arrest disparities between non-Hispanic Black and white people since Hispanic people, who could have higher arrest rates than white people, are generally coded as white.2

Still, despite the limits of this data, some clear patterns can be identified. Black people were arrested and killed by police at higher rates than white people in 34 of the 37 largest U.S. cities.

Which police departments should the feds investigate?

Arrests per 100 residents (2019) and police killings per 100,00 residents (2013-20) by race alongside disparities between those numbers for the police departments with the 37 largest jurisdictions in the U.S.

Arrests/100 Killings/100k
Police department
Albuquerque, NM 4.0 10.4 2.6 5.5 19.5 3.6
Austin, TX 2.5 9.4 3.8 4.0 7.2 1.8
Baltimore, MD 1.9 5.5 2.9 2.4 7.6 3.2
Boston, MA 0.8 2.4 2.9 0.3 5.8 17.6
Charlotte-Mecklenburg, NC 1.0 4.8 4.8 1.0 3.7 3.7
Chicago, IL* 1.7 6.8 4.1 0.3 7.4 22.1
Columbus, OH 1.0 2.5 2.6 2.5 12.7 5.1
Dallas, TX 2.0 5.0 2.5 3.1 5.1 1.6
Denver, CO 3.6 11.0 3.1 3.0 8.0 2.7
Detroit, MI 1.1 2.0 1.8 1.4 2.5 1.7
El Paso, TX 2.6 5.2 2.0 5.6 8.7 1.6
Fort Worth, TX 1.8 4.1 2.3 1.8 5.7 3.2
Fresno, CA** 5.6 11.2 2.0 3.5 2.7 0.8
Honolulu, HI 2.2 5.0 2.2 2.2 0.0 0.0
Houston, TX 1.1 3.5 3.2 1.6 7.5 4.7
Indianapolis, IN 2.8 6.1 2.2 2.1 7.5 3.5
Jacksonville, FL* 2.4 6.1 2.6 4.2 8.6 2.1
Las Vegas Metro, NV 3.9 13.2 3.4 3.3 5.9 1.8
Los Angeles, CA 1.8 4.4 2.4 1.8 8.2 4.6
Louisville Metro, KY 4.3 10.0 2.3 2.5 9.1 3.7
Memphis, TN 2.3 6.3 2.7 2.4 3.6 1.5
Mesa, AZ 3.1 12.9 4.2 4.6 0.0 0.0
Milwaukee, WI 1.2 4.4 3.8 1.0 7.0 7.3
Nashville Metropolitan, TN 2.7 6.5 2.4 0.8 3.8 4.7
New York, NY* 2.0 5.5 2.7 0.4 2.9 7.9
Oklahoma City, OK 2.1 6.3 3.0 5.0 27.2 5.5
Philadelphia, PA** 2.3 4.6 2.0 0.5 3.9 7.0
Phoenix, AZ 3.5 10.6 3.0 7.2 15.2 2.1
Portland, OR 3.0 12.8 4.3 2.9 11.1 3.9
Sacramento, CA 3.0 8.3 2.8 3.7 9.3 2.5
San Antonio, TX 2.5 9.3 3.7 3.0 10.5 3.5
San Diego, CA 2.8 8.7 3.2 2.0 3.5 1.7
San Francisco, CA 2.0 11.9 5.8 1.4 11.5 8.1
San Jose, CA 2.8 6.7 2.4 2.6 3.4 1.3
Seattle, WA 1.1 7.0 6.1 2.2 12.4 5.7
Tucson, AZ 7.2 20.2 2.8 4.2 7.9 1.9
Washington, D.C.* 0.9 6.4 7.3 0.4 5.4 13.4

*The departments serving Chicago, Jacksonville, New York and Washington, D.C., do not report their arrests disaggregated by race to the FBI, but release their data independently. We excluded arrests for traffic violations to make their data comparable to that released by the FBI.

**Data from the Fresno and Philadelphia police departments are from 2018.

Sources: Mapping Police Violence, FBI Uniform Crime Report, U.S. Census Bureau

Washington, D.C., San Francisco, Chicago and Milwaukee had some of the largest disparities in policing outcomes between Black and white residents. In these cities, Black residents were policed at high rates while white residents were policed at relatively low rates. Police arrested Black people at several times the rate of white people, even for offenses like drug possession which have been found to be committed at similar rates by Black and white communities. And police in these cities also killed Black people at substantially higher rates than white people, even after accounting for racial differences in arrest rates.

For example, the Metropolitan police in D.C. arrested Black people at a per-capita rate seven times higher and killed Black people at a rate 13 times higher than white people. In Chicago, police arrested and killed Black people at a per-capita rate four and 22 times higher, respectively, than white people. These cities have outcomes more consistent with a “pattern or practice” of discriminatory policing.

[Related: Storming The U.S. Capitol Was About Maintaining White Power In America]

But another group of cities had the highest rates of arrests and killings by police overall. Police departments in Southwestern cities like Tucson, Phoenix and Albuquerque arrested and killed both white and Black people at relatively high rates. And Oklahoma City police had the highest rate of police killings overall. These outcomes may reflect a broader pattern of unlawful or excessive use of police powers.

These initial findings are not enough to prove policing in these cities is unlawful or to prove these disparities were motivated by racial bias. But they do establish a set of “warning flags” that signify undesirable policing outcomes that, at minimum, should result in further inquiry and investigation from those tasked with holding police accountable. Indeed, the Justice Department had already flagged many of these cities for reform during the Obama administration — including in Albuquerque, San Francisco and Chicago. Many of these reforms are still in the process of being implemented. Other cities — Tucson, Oklahoma City and Phoenix — have not been investigated or reformed by the federal government. These cities might benefit most from Department of Justice intervention.

So what should the Biden administration do with this information? It could start by opening DOJ investigations into police departments, like those in Phoenix and Oklahoma City, that have largely evaded federal intervention despite reporting relatively high rates of police violence. Now that Democrats have taken control of Congress, they could give this administration additional tools and resources to conduct these investigations. The police reform bill sponsored by Senate Democrats, the George Floyd Justice in Policing Act, would empower the Justice Department to collect additional data from local police departments on nonfatal force, stops and searches, and misconduct allegations to better identify places where interventions might be needed. These data could inform a more expansive analysis that accounts for additional factors such as calls for service, crime rates and other contextual factors.

But finding places that have the most problematic policing practices is not the only problem. Another problem is one of scale — the Justice Department has opened only 69 formal “pattern or practice” investigations since 1997. To change policing in a nation with over 18,000 police departments, the Biden administration would need to dramatically increase these interventions, as well as move beyond the largest cities and into rural and suburban areas where police violence is increasing. Whether or not the Biden administration will end up taking this approach to scale remains to be seen.

Why police aggression is far more pronounced against left-leaning protesters

Confidence Interval: QAnon is not going anywhere | FiveThirtyEight


  1. Some police departments serve consolidated city-counties, or their jurisdictions include a city and its surrounding county (e.g., the Las Vegas Metropolitan Police Department and the Charlotte-Mecklenburg Police Department). In those situations, I used census data at the county level, as that fully captures the department’s jurisdiction. Any police killing that involved an officer of a given police department, according to the Mapping Police Violence database, is included.

  2. Police departments in four cities — Chicago, New York, Jacksonville and Washington, D.C. — did not report any arrests data disaggregated by race to the UCR. Arrests data was obtained from municipal and state databases instead. We excluded arrests for traffic violations, which are not included in the federal UCR database, to ensure consistency among agencies.

Samuel Sinyangwe is a data scientist, policy analyst and co-founder of Campaign Zero.