Which sources track police killings by race and how do their counts differ?
Executive summary
Major independent trackers of U.S. police killings include The Washington Post’s Fatal Force database, Mapping Police Violence, academic studies that compile multiple sources, and aggregated sites such as World Population Review and university projects; The Washington Post reports about “more than 1,000” people shot and killed per year on average and provides detailed race-by-case counts since 2015 [1]. Mapping Police Violence includes broader categories of police-caused deaths (not just shootings) and reports race and census-based rates that typically show higher per-capita rates for Black and Hispanic people than for whites [2] [3].
1. Who is counting and what each tracker includes
The Washington Post’s Fatal Force project logs every person shot and killed by on-duty police officers from Jan. 1, 2015 through Dec. 31, 2024 using local reporting, law‑enforcement releases, social media and other databases; it focuses on fatal shootings and collects dozens of case details including race [1]. Mapping Police Violence compiles a broader set of police-caused deaths — shootings plus deaths from restraint, vehicle collisions and other actions — and pairs incident data with 2020 census race/ethnicity denominators to compute rates [2]. Academic groups and projects — for example, Johns Hopkins/Bloomberg School studies and the Law Enforcement Epidemiology Project at UIC — analyze compiled incident datasets to estimate injuries, fatalities and racial disparities, sometimes combining fatal and nonfatal shootings to capture a larger burden [4] [5].
2. Why counts differ: inclusion rules and methods
Counts diverge because trackers define the universe differently. The Washington Post counts fatal shootings by on‑duty officers and excludes other police-related fatalities unless shot; Mapping Police Violence includes non‑shooting deaths such as suffocation or restraint and therefore typically reports larger totals for “police killings” [1] [2]. Academic analyses may merge multiple sources, include nonfatal injuries, or apply epidemiological methods to estimate undercounted events, producing different totals and denominators [4] [5].
3. Race reporting: raw counts versus rate-per-capita
Raw counts often show more white people killed numerically because white Americans are the largest racial group in the U.S., but per-capita rates tell a different story: The Washington Post’s rate calculations show Black Americans face higher fatal‑shooting rates (for example, roughly 6.2 per million vs. 2.4 per million for whites in 2015–2024 reporting cited) [6] [1]. Mapping Police Violence’s use of census denominators similarly highlights elevated per‑capita risk for Black and Hispanic communities and also documents disparities in circumstances [2] [7].
4. Data gaps and reporting biases that shape racial totals
All trackers note incomplete or inconsistent official reporting from police departments; The Washington Post’s project began because FBI reporting undercounted shootings, and it continues to fill gaps by mining media and other databases [1]. World Population Review and other aggregators warn that “unknown” race entries and inconsistent local data increase uncertainty in per‑race totals and can shift apparent trends year to year [3]. Academic work emphasizes that fatal counts alone understate the burden when nonfatal injuries are excluded, and race is sometimes missing in nonfatal datasets [4] [5].
5. What research finds when you combine methods
Peer‑reviewed studies and university projects that merge incident datasets or study both fatal and nonfatal shootings find larger racial disparities than single‑source death counts imply: Johns Hopkins and Vanderbilt researchers estimated 1,769 police‑shooting injuries per year in 2015–2020 (55% fatal) and found Black non‑Hispanic people were overrepresented among nonfatal injuries [4]. Other academic work links county‑level factors such as income inequality to higher fatal‑shooting counts, with Black and Hispanic communities bearing greater risk [8].
6. How to read headline numbers responsibly
Ask which incidents are counted (shootings only versus all police‑caused deaths), whether numbers are raw counts or per‑capita rates, how race was assigned or recorded, and whether nonfatal harm is included. The Washington Post provides case‑level transparency for fatal shootings since 2015 [1]; Mapping Police Violence expands scope to capture deaths beyond shootings and uses census denominators [2]. Academic analyses add methodological adjustments and context about nonfatal injuries and structural correlates [4] [8].
Limitations: available sources do not mention every possible tracker (for example, Fatal Encounters or official FBI Crowdsource projects are referenced historically but not detailed here), and the datasets cited cover overlapping but non‑identical time windows and inclusion rules [1] [2] [4]. For any specific year or racial breakdown you need, consult each tracker’s methodology page and compare raw counts and per‑capita rates side‑by‑side using the primary source datasets [1] [2] [6].