How have homicide victimization rates per 100,000 varied by race in the U.S. over the last 30 years?
Executive summary
Over the last three decades homicide victimization rates in the United States have fallen from the high levels of the early 1990s, but racial disparities have remained large and persistent: Black and American Indian/Alaska Native (AI/AN) populations have consistently faced substantially higher homicide rates per 100,000 than White, Hispanic, and Asian groups [1]. A sharp national rise around 2019–2020 interrupted long declines, but the uneven burden by race continued through 2023 with Black victims experiencing rates many times higher than White victims [2] [3].
1. Long-term national trajectory: decline, spike, partial rebound
After very high homicide rates in the late 1980s and early 1990s, overall U.S. homicide rates trended downward into the 2010s, a pattern documented in long-run BJS trend analyses and CDC data, only to see a notable single-year jump between 2019 and 2020—the largest in over a century—and elevated levels through 2021 before some decline by 2024–2025 [2] [4].
2. The size of racial disparities: multiples, not marginal differences
Across the last thirty years the gap between Black and White homicide victimization rates has not simply persisted but often been large in magnitude: in recent years Black homicide rates have been roughly four to seven times the White rate depending on year and dataset (for example, 2023 figures showing ~26.6 per 100,000 for Black victims versus ~3.9 for White victims in CDC-based tabulations) [3] [5].
3. Which groups carry the highest burden: Black and AI/AN men peak highest
Disaggregated studies show that American Indian and Alaska Native and Black males—particularly ages roughly 15–44—have the highest homicide rates of any demographic groups; JAMA/GBD analyses through 2019 and CDC/BJS breakdowns show these concentrated, age-sex-specific peaks [1] [6]. Women of color, especially Indigenous women, also face higher homicide victimization rates than White or some Hispanic female counterparts [7] [1].
4. Geography and concentration: place amplifies race differences
Much of the racial gap is concentrated in particular places—cities and counties with entrenched poverty, segregation and inequality—and county-level GBD/IHME estimates demonstrate substantial within-race geographic variation that amplifies overall disparities [8] [6]. State-level reports show some states (e.g., Missouri in recent CDC tabulations) with especially high Black homicide victimization rates, illustrating how local context shapes national racial patterns [5].
5. Why disparities persist: structural and proximate explanations
Research attributes racial differences to a mix of proximate dynamics—most homicides are intraracial, and young males drive rates—and structural causes including concentrated poverty, inequality, housing segregation, and differential exposure to firearms; academic studies and policy analyses emphasize socioeconomic and neighborhood mechanisms rather than simple individual pathology [9] [10] [4].
6. Recent years (2019–2023): spike, persistence, and weapon patterns
The 2019–2020 increase in homicide disproportionately affected certain racial groups and younger males, and recent 2023 CDC-based analyses and advocacy reports found that Black victims continued to make up a disproportionate share of homicides (over half in some counts) and that firearm homicides were a larger share of Black victimizations compared with White victims [4] [3] [5]. Commentary from Council on Criminal Justice and BJS underscores that overall violent crime trends fell again by 2024–2025 but racial disparities in victimization rates remained stubborn [2].
7. Data strengths, gaps and caution in interpretation
National mortality systems (CDC/NCHS), BJS surveys, and modeled county estimates (GBD/IHME) converge on the broad picture of persistent disparities, but differences in race/ethnicity classification, undercounting in some jurisdictions, and the modeling horizon (many granular datasets stop at 2019) mean year-to-year comparisons and exact ratios vary depending on source; available sources do not permit definitive attribution of every recent fluctuation to single causes [11] [8].