World Cup visitors discovered something many Americans already know: The United States does not feel like the dangerous place they’ve been led to expect. That’s because America’s murder problem is not spread evenly across the country. Does the United States have a higher murder rate than its peer countries? Yes – but new research shows that statistic obscures how murders in the U.S. are extremely concentrated in very tiny areas.
In 2025, the U.S. murder rate fell to just under 4 murders per 100,000 people. During the first six months of 2026, data compiled by the Council on Criminal Justice shows that murders in the nation’s 30 largest cities declined by another 18%. If that decline reflects the national trend, and that is likely, the U.S. murder rate will fall to approximately 3.25 per 100,000 people this year.
Even at that level, the U.S. murder rate would remain about 70% higher than Canada’s 2024 homicide rate of 1.91 per 100,000 people and Australia’s approximately 2 per 100,000.
Nevertheless, the United States has recently made remarkable progress. Since national recordkeeping began in 1900, the country has never recorded a murder rate this low. The previous record low was approximately 4.5 murders per 100,000 people, and the rate stood at about 5.0 per 100,000 in 2024.
At the same time, focusing solely on the national murder rate paints an incomplete picture. Murders in the United States are extraordinarily concentrated in a small number of places, while the nation’s overall violent crime victimization rate remains below that of many other developed countries.
The research by the Crime Prevention Research Center, which I head, finds that 31 counties with the highest murder rates – just 1% of all U.S. counties – contain about 20% of the nation’s population but account for 41% of all murders. Expanding the analysis to the 63 highest-rate counties (2% of counties) captures 30% of the population but 54% of all murders.
The concentration does not stop at the county level. Within these counties, murders cluster into remarkably small geographic areas. Roughly two-thirds of all murders in the worst 2% of the counties occur within about ten-block neighborhoods, whose populations represent only a tiny fraction of their counties’ residents. Much of this violence is associated with street gangs and repeat offenders rather than random acts committed throughout the broader community.
Most of the country experiences very little murder. Fifty-three percent of U.S. counties record no murders in a typical year, and another 16% record only one murder. Together, these 69% of counties account for just 3% of all murders nationwide.
Media coverage often leaves the impression that murderers are ordinary people who suddenly lose their tempers. The evidence tells a very different story. Murder offenders overwhelmingly have extensive criminal histories. About 90% of murderers already have criminal records, and some portion of the remaining 10% almost certainly committed crimes before the homicide but had not yet been caught.
Criminologists have long recognized that violent crime is committed by a relatively small group of chronic offenders. Research has consistently shown that 5% of the criminals account for most crime. Marvin Wolfgang’s landmark Philadelphia birth cohort studies found that roughly 6-8% of offenders accounted for the majority of serious violent crime, including about 71% of murders, 73% of rapes, 82% of robberies, and 69% of aggravated assaults. Subsequent research has repeatedly confirmed that a small fraction of chronic offenders commit a disproportionate share of violent crime.
Who commits murders also differs dramatically across demographic groups. In 2020, black men made up about 6.0% of the U.S. population but accounted for 49.7% of known murder offenders – about 8.3 times their share of the population. Black women accounted for 5.5% of murder offenders while making up 6.43% of the population, or about 0.85 times their population share.
By comparison, white men (including Hispanics) comprised 30.3% of the population but 35.7% of murder offenders. White women accounted for just 6.1% of murder offenders despite making up 31.3% of the U.S. population. Asian men and women had the lowest rates of murder offending, accounting for only 0.43 times and 0.06 times, respectively, their shares of the U.S. population.
Many look at these numbers and say this proves the criminal justice system is racist. Many Democrats think the share of blacks in prison shouldn’t exceed their share of the population (see also here and here). For example, in 2020, the Democrats on the Congressional Joint Economic Committee complained, “Black Americans made up 33% of the sentenced prison population in 2017, despite comprising only about 13% of the U.S. adult population. The vast disparity in incarceration rates is driven substantially by the unequal application of U.S. laws.”
The problem is that these discussions ignore the race of the victims. Ninety percent of black murder victims are murdered by other blacks. If you make it so there is little or no penalty for murdering blacks, you ensure that more blacks will be murdered.
The lesson is straightforward. Although the United States has a higher national murder rate than many peer countries, murders are extraordinarily concentrated among a small number of repeat offenders operating in a tiny number of neighborhoods. If you avoid hanging around people with violent criminal records and those places, you are very unlikely to become a victim.
Yet many Democrats draw the wrong conclusion from these data and advocate policies that will increase the number of black murder victims.
