Assault victims in 59 large US cities

In all 46 large US cities with complete police data for 2022 to 2025, Black women's reported assault rate is higher than White women's (about 3.8 times in the typical city) and, in all 41 that record ethnicity, higher than Hispanic women's (about 2.4 times). Nothing measured here explains why.

Women only (749,269 Black women, 668,499 other women). Each chart tests one plain explanation.

2022 to 2025

2024 to 2025 only

Women victims of aggravated and simple assault per 100,000 residents of each group a year, from each city's own police department in the FBI's NIBRS files; Hispanic of any race first. Population: ACS 2020 to 2024 via Census Reporter. Built with disparity-kit using the Dallas decisions. Built with disparity-kit.

• What, not why. Shows how often assaults are reported, not why. Nothing here measures causes or offenders.
• Police reports only. Willingness to report, and where police patrol, differ by group, place and time. This data cannot separate them from differences in assaults.
• Reports, not people. Someone assaulted twice counts twice.
• No neighborhood test here. The FBI files have no victim location. Where neighborhoods were tested (Los Angeles, Baltimore, Dallas; New York's precincts in the follow-up), the gap narrowed but remained: about 3x within New York precincts of like composition.
• Levels not comparable across cities. Departments code offenses differently: intimidation, left out, is 0.1% of assault-type victims in Newark and 51% in New York. Compare ratios within a city, not rates across cities. Columbus changed its coding in November 2024; Tucson's 2025 looks incomplete.
• Race recorded by officers. The Census counts Black alone; 3% to 56% more residents are Black alone or in combination, which sets the lowest bound.
• Ethnicity. Five cities record none (Toledo, Cleveland, Tulsa, Detroit and Oklahoma City): no Hispanic comparison, and the ratio to White women is a floor. Three code only Hispanic (Portland, Raleigh and Columbus).
• Unknown race. Up to 23% of women victims (Seattle). If every one were in the comparison group, the gap with Hispanic women would vanish in Seattle and Boston; the gap with White women stays above 2x.
• Intervals show counting noise only. They leave out repeat victimization, coding error and the bounds above.
• Where people live. Rates divide by residents, not where people spend time. A department serving more people than its city (Charlotte, 1.15x) runs high.
• Small numbers. Asian women's counts are small, so their ratios stay out of the headline; fewer than 10,000 Hispanic women live in Cincinnati, St. Louis and Chesapeake.
• Choices made after seeing the data. The screen's thresholds, the typical-city medians and keeping Columbus and Tucson. All are disclosed; none were set in advance.
• One department each. Transit, school, campus and county police are not counted.
• Not national. Large cities with complete data, not a national sample.