“Black-on-Black crime” turns murdered Black people into evidence against Black America.
Section 1 of 9 · Introduction
“Black-on-Black crime” turns murdered Black people into evidence against Black America.
Real victims, same-race violence and the attempt to turn a crime statistic into an answer to police accountability.
By Al Agami · XORCIS.AI · 1 October 2026

Black Americans bear a disproportionate burden of homicide. The Bureau of Justice Statistics estimated 11,060 Black homicide victims in 2023, with a victimization rate of 21.3 per 100,000 people, compared with 3.2 for White people. These are victims, not a count of criminals. [4]
The same-race pattern is also real. In the FBI’s selected 2019 single-victim, single-offender records, Black offenders were recorded for 2,574 of 2,906 Black victims; White offenders for 2,594 of 3,299 White victims. Those shares are 88.6% and 78.6%, retaining unknown offender race in each denominator. [1]
In a clip broadcast by CNN on 11 July 2016, Rudy Giuliani invoked killings of Black people by other Black people to attack Black Lives Matter’s focus on police. He also accused the movement of endangering officers. The transcript records his argument; it does not verify every statistic or accusation in the clip. [5]
Turning murdered Black people into evidence against Black America is this article’s interpretation of the broadcast argument, and the title’s forensic satire. A victim becomes an exhibit against a racial population’s demands for protection. That collective accusation is a rhetorical move, not a finding established by the homicide records.
A description of victim and offender can be accurate. An explanation of the crime requires more.
“Black-on-Black” tells the reader that two recorded racial categories match. The label supplies no motive, identifies no weapon, establishes no relationship and proves no racial cause. A racial pairing is not a mechanism.
The rhetorical invention arrives when that pairing becomes a verdict on a population, or an instruction to stop asking about state violence. A homicide table cannot issue either ruling.
Three questions need separate answers: who harmed the victim; how frequently a population suffers harm; and whether a particular police action was lawful.
A percentage of victims killed by same-race offenders answers a different question from a population crime rate. It does not mean that the same percentage of Black Americans commits murder. Moving from victims in selected cases to an entire population changes the denominator and the claim.
Likewise, more civilian murders do not settle the evidence in a police killing. Even a correctly counted larger category cannot acquit an officer in a different case. Investigation of civilian violence and scrutiny of public authority can proceed together.
Relationships, opportunities for contact, local conditions and access to weapons are questions for evidence. The racial matrix alone cannot separate those explanations. Treating a category as a cause simply skips the investigation.
The White-victim comparison defeats the suggestion that same-race homicide is exclusively a Black phenomenon. It does not make the percentages identical, erase unequal homicide risk or excuse a single offender.
A separate BJS survey found perceived same-race offenders in 70.3% of nonfatal violent incidents involving Black victims and 62.1% involving White victims in 2018. Table 14 excludes incidents with unknown perceived offender race or ethnicity; those were 11% of violent incidents overall. White and Black categories there exclude Hispanic people. These survey estimates concern people aged 12 or older, include crimes not reported to police, and are not homicide or conviction statistics. [3]
Nor is the FBI chart a complete murder ledger. Table 6 requires one victim, one offender and some known offender information. The broader Table 4 separately lists 3,533 single-victim records with an unknown offender or offenders. Unknown cases cannot silently acquire a race. The FBI records race and Hispanic ethnicity separately, so its racial categories are not identical to the survey’s. [2] [1]
Concern about violence within Black communities is legitimate. Effective prevention, support for survivors and accountability for perpetrators remain urgent. None requires collective guilt.
The records establish substantial same-race violence in the specified datasets, and a severe, unequal burden of homicide. The 2018 and 2019 comparisons are deliberately dated historical evidence, not claims about the latest year.
The records do not establish an innate racial tendency, prove that Black communities ignore victims, or decide whether a police officer used lawful force. Giuliani’s broadcast argument is evidence that the comparison was deployed politically, not proof of the accusations attached to it.
Individual responsibility needs individual evidence. A victim’s race cannot transfer responsibility to millions of strangers. A murderer’s race cannot cancel a public institution’s duty to account for its own conduct.
- What is being counted: homicide records, nonfatal survey incidents, arrests or convictions, and which unknown cases are excluded?
- What does the comparison establish after keeping the year, racial definitions and denominator consistent?
- What would protect victims and hold each offender accountable, including an officer when the evidence warrants it?
Forensic satire: The victim has been entered as evidence against the victim’s race. The murderer must be relieved to have so many co-defendants.
Sources
- [1] FBI, Expanded Homicide Data Table 6, 2019. Selected victim–offender records; chart counts and exclusions.
- [2] FBI, Expanded Homicide Data Table 4, 2019. Broader victim/offender situations, including unknown offenders.
- [3] BJS, Criminal Victimization, 2018, September 2019, Table 14, page 13 and methodology. Nonfatal survey incidents, perceived offender characteristics and category definitions.
- [4] BJS, Homicide Victimization in the United States, 2023, May 2025, revised 19 August 2025, Table 1. National estimates using imputation and weighting for missing data and nonresponse; homicide means murder and nonnegligent manslaughter.
- [5] CNN, Anderson Cooper 360°, 11 July 2016, rush transcript. Broadcast clip of Giuliani and panel responses. Used to document the argument, not to adopt unverified claims by participants.