What Does ‘13/51’ Actually Mean?

Abstract

The expression “13/51,” sometimes rendered “13/50,” claims that Black Americans are roughly 13% of the United States population but commit about half of crime. This brief reconstructs the claim from the primary federal tables most often used to support it and tests each substitution required to turn those tables into the slogan.

The 51.2% value appears in the FBI’s 2019 Table 43 as the Black share of arrests for murder and nonnegligent manslaughter among 10,831 reporting agencies covering an estimated population of 229,735,355. In the same table, Black people accounted for 36.4% of violent-crime arrests and 26.6% of all arrests. A separate supplemental homicide table reported that Black offenders were 55.9% of known-race offender records, but race was unknown in 4,752 of 16,245 records.

The evidence establishes a severe racial disparity in homicide arrest, offending, and victimization measures. It does not establish that Black Americans commit half of all crime, that half of Black people offend, or that race itself causes the disparity.

Key findings

Claim reconstruction

Where does the “51” come from?

The cleanest documented source is the FBI’s Crime in the United States, 2019, Table 43. Participating agencies reported 7,964 arrests for murder and nonnegligent manslaughter. Of those arrest events, 4,078 were recorded as Black or African American—51.2%.

That sentence is accurate only when every limiting phrase remains attached: murder arrests, reported by participating agencies, in 2019. Remove those limits and the statistic changes meaning.

Measure comparison

The percentage changes when the question changes.

The denominator problem

“Thirteen percent” is not an exposure-adjusted comparison.

The population side of the slogan uses the entire Black population: children, elders, women, men, people in low-crime communities, and people in high-violence neighborhoods. The arrest side is concentrated by age, sex, place, and exposure to violence. Comparing one group’s share of the total population directly with its share of homicide arrests describes disproportionality, but it does not adjust for the population actually at risk of arrest or explain why the disparity exists.

Race categories also require care. “Black alone,” “Black alone or in combination,” and “non-Hispanic Black” are different Census denominators. FBI race and ethnicity fields are collected separately and ethnicity reporting is incomplete. A viral two-number ratio usually suppresses those classification choices.

Arrests are not all crime

Law-enforcement data measure contact with a reporting system.

An arrest table records arrests made by law enforcement. It does not count offenses never reported to police, offenses recorded without an arrest, or every person responsible for an offense. Arrest totals can also include the same individual more than once and do not establish conviction.

The Bureau of Justice Statistics therefore treats the FBI’s incident-based records and the National Crime Victimization Survey as complementary rather than interchangeable. The NCVS interviews people about nonfatal victimizations reported and not reported to police; NIBRS records crimes known to participating law-enforcement agencies. In 2023, for example, BJS estimated that 42% of robbery victimizations were reported to police. No single source is a census of every crime committed.

Known and unknown offenders

The 55.9% figure is conditional on known race.

The FBI’s separate 2019 Expanded Homicide Data Table 3 listed 16,245 offender records. Race was recorded as Black for 6,425, White for 4,728, another race for 340, and unknown for 4,752.

Neither percentage is inherently fraudulent; they answer different questions. Responsible writing must state whether unknown records remain in the denominator.

The missing half of the story

Black people are not only counted as suspects. We are disproportionately counted among the dead.

In the FBI’s 2019 supplemental homicide data, 7,484 Black people were recorded among 13,927 murder victims. Black people were 54.7% of victims whose race was known. Among the single-victim/single-offender cases in Table 6, 2,574 of 2,906 Black victims were killed by a Black offender, while 2,594 of 3,299 White victims were killed by a White offender. Because the table excludes cases where all offender information was unknown, it should not be generalized to every homicide.

CDC’s National Violent Death Reporting System found that Black people accounted for 58.9% of male homicide victims in 2022. Black males had the highest homicide victimization rate, 56.0 deaths per 100,000. The honest correction to “13/51” cannot become denial: concentrated lethal violence is a major Black public-safety and public-health emergency.

The dead cannot be reduced to ammunition in a racial argument. The measure must lead back to protection, prevention, and Black life.

What the data cannot explain alone

A disparity is not a causal theory.

These tables are descriptive. They do not isolate the effects of age structure, sex, firearm exposure, concentrated poverty, residential segregation, neighborhood instability, illegal-market activity, policing patterns, clearance practices, or local institutional capacity. They cannot establish a biological or cultural cause from a racial label.

CDC’s 2022 violent-death review points to concentrated poverty, stressed local economies, residential instability, neighborhood disorganization, and low community cohesion as conditions associated with unequal exposure to violence. Testing causal explanations requires designs that measure those conditions; a national race percentage cannot substitute for that work.

Why one year cannot become a permanent identity

The slogan freezes 2019 while the data keep moving.

The FBI changed its national reporting architecture after 2019, expanded NIBRS participation, and now releases data through the Crime Data Explorer. In 2024, more than 16,000 agencies covering 95.6% of the U.S. population submitted data, and the estimated number of murders fell 14.9% from 2023.

That national decline does not by itself reveal how every racial disparity changed. It does show why a single pre-pandemic table should not be treated as a timeless racial law. Updated claims require updated, definition-matched data.

Data and method

A claim audit readers can reproduce.

1. Reconstruct the claim: Identify the primary table and exact row producing each number.
2. Name the unit: Distinguish arrests, offenses, offender records, victims, convictions, and unique people.
3. Audit coverage: Record year, participating agencies, population coverage, and reporting system.
4. Inspect missingness: Recalculate percentages with and without unknown-race records.
5. Compare adjacent measures: Test whether the result survives movement from murder arrests to violent-crime and all-arrest categories.
6. Restore victimization: Examine who is harmed, not only who is arrested or identified as an offender.
7. Separate description from cause: Do not infer why a disparity exists from a descriptive race table.

Limitations

• This brief audits the most common 2019 formulation; online variants may use a different year or table.
• Arrest data reflect crimes known to police and enforcement practices, not a complete census of offending.
• Supplemental homicide records contain substantial missing offender information.
• Race and ethnicity definitions differ across Census, FBI, BJS, and CDC systems.
• National aggregates conceal enormous variation by age, sex, city, neighborhood, and time.
• The descriptive data cannot identify the causal contribution of structural, neighborhood, institutional, or behavioral conditions.
• Newer FBI systems should not be joined mechanically to the 2019 legacy table without a comparability review.

Conclusion

The slogan is wrong. The crisis is real.

“Thirteen percent commits fifty-one percent of crime” is not what the FBI measured. A defensible sentence is narrower: among arrest events reported by 10,831 agencies in 2019, Black people accounted for 51.2% of murder arrests. In that same table, the Black share was 36.4% for violent-crime arrests and 26.6% for all arrests.

Correcting the statistic should not become an excuse to avert our eyes from homicide. Black people are disproportionately represented among both homicide offenders and victims, and Black male victimization is especially severe. A pro-Black response refuses both racial fatalism and statistical denial. It demands exact measurement, prevention, accountable public institutions, community capacity, and the protection of Black life.