Firearm Homicide Surges From 1979 to 2023
ResearchPosted on rand.org Nov 20, 2026Published in: JAMA Network Open, Volume 9, No. 11 (November 2026). DOI: 10.1001/jamanetworkopen.2026.44022
ResearchPosted on rand.org Nov 20, 2026Published in: JAMA Network Open, Volume 9, No. 11 (November 2026). DOI: 10.1001/jamanetworkopen.2026.44022
US firearm homicide rates surged twice in the past 35 years. Better understanding of such surges could aid prevention.
To compare the demographic and geographic distribution of surge deaths and examine state-level correlates of surge severity across the 1985-1993 and 2014-2021 firearm homicide surge periods.
This serial cross-sectional study obtained homicide counts from 1979 to 2023, stratified by race, sex, age group, county urbanicity, and state, from the Centers for Disease Control and Prevention. The firearm homicide surge periods 1985 to 1993 and 2014 to 2021 were analyzed. Although both Black and Hispanic people in the US bear a disproportionate burden of firearm homicide, Hispanic ethnicity was not consistently coded in mortality records over the study period (1979-2023); thus, a single categorization of Black or other race was used. Data analyses were conducted in April 2025 to September 2026.
Presurge levels and presurge-to-surge changes in state measures of drug-market violence, proxy measures for firearm availability and use, economic strength, social or economic disadvantage, urban concentration, social or economic disruption, and policing capacity.
Annualized firearm homicide surge rates (deaths per 100 000 population, above presurge rates). Correlations between state-level surge rates and proxy measures were calculated.
The 2014-2021 surge produced an estimated mean (SE) of 31 489 (509) firearm surge deaths, with a mean (SE) annualized rate of 1.20 (0.02) deaths per 100 000 population per year above presurge rates. In the 1985-1993 surge, the mean (SE) number of firearm surge deaths was 24 276 (580), with a mean (SE) annualized rate of 1.09 (0.03) deaths per 100 000 population. Both surges were driven entirely by firearm homicide, as non–firearm homicides declined over both periods. Deaths of young urban Black males (aged 15-24 years) accounted for 47% of 1985-1993 surge deaths but only 14% of the 2014-2021 surge. Compared with the 1985-1993 surge, during the 2014-2021 surge, the mean (SE) firearm homicide surge rates increased for females (0.15 [0.02] vs 0.38 [0.01]), people of races other than Black (0.22 [0.02] vs 0.48 [0.01]), and rural residents (0.13 [0.04] vs 0.83 [0.03]). Across US states, the 1985-1993 surge correlated with crack cocaine market activity (r = 0.44; 95% CI, 0.19-0.64), while the 2014-2021 surge correlated with presurge firearm availability proxies (r = 0.68 [95% CI, 0.50-0.81] for value of stolen firearms per capita) and COVID-19–related death rates (r = 0.61; 95% CI, 0.40-0.76). Across both eras, higher state-level surge rates correlated with social or economic disadvantage, including premature mortality, poverty rates, lower employment, and household instability.
In this serial cross-sectional study of US firearm homicide mortality, both the 2014-2021 surge, which was substantially larger and more demographically and geographically dispersed, and the 1985-1993 surge were associated with distinct proximal correlates but disproportionately affected states with broader structural vulnerabilities. Proximal drivers of rapid increases in homicide rates may exacerbate long-standing population vulnerabilities.
This publication is part of the RAND external publication series. Many RAND studies are published in peer-reviewed scholarly journals, as chapters in commercial books, or as documents published by other organizations.
RAND is a nonprofit institution that helps improve policy and decisionmaking through research and analysis. RAND's publications do not necessarily reflect the opinions of its research clients and sponsors.