How do police shooting rates by race compare when adjusted for crime rates and population?

1. The headline numbers: per‑capita gaps are large and well‑documented

Multiple databases and aggregators show higher per‑person fatal‑shooting rates for Black Americans than for white Americans — Washington Post–derived charts summarized on Statista report roughly 6.1–6.2 deaths per million for Black people versus about 2.4 per million for white people across 2015–Dec. 2024 ^{[2] [1]}. These per‑capita comparisons underpin much public concern and policy debate ^[2].

2. Why researchers adjust for crime rates — and how that changes the story

Critics of simple per‑capita comparisons argue police shootings should be benchmarked to the relevant exposure: local violent crime rates, arrests, or police contact rates. Studies that re‑weight shootings against measures of criminal involvement sometimes reduce or eliminate measured racial disparities, leading some scholars to conclude differences reflect differential exposure to violent crime rather than officer bias ^{[8] [6]}. Proponents of the benchmark approach say it better matches the contexts in which shootings occur ^{[9] [8]}.

3. Evidence that disparities persist after adjusting for crime and place

Major reviews and independent analyses find that fatal police violence remains disproportionately high for Black people even when accounting for crime, place, and socioeconomic vulnerability. A Lancet network meta‑regression concluded police are more likely to shoot Black civilians than White civilians given the same levels of criminal activity ^[3]. Other studies show social‑vulnerability and neighborhood composition strongly predict fatal shooting rates — with larger increases for Black and Hispanic residents in high‑SVI (Social Vulnerability Index) areas ^{[10] [11]}.

4. Studies that find no systematic anti‑Black disparity after benchmarking — and their critics

Some influential papers report no systematic anti‑Black disparity once shootings are benchmarked to race‑specific crime rates (Cesario, Johnson & Terrill; an earlier PNAS article that was later retracted) ^{[6] [7]}. These results provoked sharp methodological pushback: critics argue crime and arrest data are imperfect proxies for exposure to police, can be endogenous to biased policing, and that different benchmark choices produce very different conclusions ^{[12] [4] [13]}.

5. Data quality, benchmarks and hidden assumptions matter more than you think

All sides acknowledge the same limitation: available datasets are incomplete and choice of benchmark drives results. Public databases undercount incidents, vital‑records misclassification exists, and crime data (UCR/NCVS) diverge substantially; using arrests or crime rates to “explain away” disparities risks circularity because policing patterns influence those benchmarks in the first place ^{[3] [14] [9]}. Scholars warn that “disparity ≠ bias” is true in principle, but benchmark methods can also mask bias if the benchmark itself is affected by biased policing ^{[8] [4]}.

6. Place, trauma care and survivability complicate fatal‑only comparisons

Analysts note that studies focused only on fatalities miss nonfatal shootings and differences in shot survivability tied to proximity to trauma centers, wound location, or first‑aid practices; racial differences in these factors can bias fatality comparisons independent of shooting incidence ^{[15] [16]}. Johns Hopkins and others show disparities when nonfatal as well as fatal injuries are analyzed ^{[16] [15]}.

7. Bottom line for readers and policymakers

Per‑capita figures show a clear racial gap in fatal police shootings ^{[2] [1]}. Whether that gap is best explained by differences in criminal‑activity exposure, policing practices, place‑based social vulnerability, or a mix of those factors is contested in the literature; major reviews and several peer‑reviewed studies conclude disparities persist after adjustment, while other analyses using different benchmarks conclude otherwise ^{[3] [6] [4]}. Given data limits and methodological disputes, the strongest policy inference is practical: improving data collection (complete, encounter‑level reporting), measuring both fatal and nonfatal shootings, and addressing concentrated disadvantage and policing practices would reduce uncertainty and help target reforms ^{[14] [10]}.

Limitations: available sources disagree on causation and differ in benchmarks and scope; this summary relies only on the supplied reporting and studies and does not assert claims beyond them.