JAMii Myth Buster Series
A Black and white image showing a child near a fence with sunlight in the background, symbolizing hope and freedom beyond harmful narratives.
Myth Buster Series

Debunking the Phrase “Black-on-Black Crime”

The violence is real. The pain is real. But the phrase is often used to make Black people look uniquely violent, while hiding the bigger story: most violence is local, same-group violence happens across communities, and arrest data is not the same as crime reality.

Data + Context Easy-Read Charts Bias Model Included Sources Listed

Myth: “Black-on-Black crime” proves Black people are uniquely violent, self-destructive, or broken.

Truth: The phrase is misleading. Same-group violence happens in every group because people are usually harmed by people near them. White victims are also most often victimized by White offenders.

The simple answer

“Black-on-Black crime” is not a neutral phrase. It is often used as a racial narrative. It focuses attention on Black pain while ignoring the conditions that create violence and the fact that same-group violence is normal across racial and ethnic groups.

First, we cleaned the list

Before looking at the numbers, the categories had to be cleaned. That matters because some crime data mixes race and ethnicity in confusing ways. For example, many Hispanic/Latino people may be counted as White in some arrest data. For this article, the cleaner Bureau of Justice Statistics (BJS) victimization categories were used.

Here is the cleaned list:

Group How this article counts it
WhiteNon-Hispanic White only.
BlackNon-Hispanic Black only.
Hispanic/LatinoListed separately. Not hidden inside White or Black.
Asian / Native Hawaiian / Pacific IslanderListed separately, but BJS warns smaller sample sizes can make this estimate less stable.
Other / AIAN / MultiracialAmerican Indian/Alaska Native and people of two or more races. Not counted as Black or White.
Unknown offender race/ethnicityRemoved from the final adjusted percentage because the offender group was not known or perceived.

Source note: BJS states that White and Black categories in this victimization table exclude Hispanic origin. BJS also states that “Other” includes American Indian/Alaska Native or persons of two or more races, and that offender race/ethnicity is based on victims’ perceptions.

Second, we used victimization data first

Victimization data is not the same as arrest data. Victimization data asks people whether they were harmed. Arrest data tells us who police arrested. That is an important difference.

Victimization data

Measures who experienced crime. The broader NCVS includes crimes reported and not reported to police. For the bias model in this article, we use CSG’s 2023 comparison of violent victimization and violent arrests, which focuses on victimization estimates tied to reported-to-police incidents.

Arrest data

Measures who was arrested. It can be affected by patrol patterns, police discretion, neighborhood surveillance, reporting, and bias.

Actual cleaned same-group numbers

After cleaning the categories and removing unknown offender race/ethnicity, the same-group violent incident numbers look like this:

Actual cleaned same-group violent incidents

BJS/NCVS violent incidents, 2017–2021. Unknown offender race/ethnicity removed.

Black
68.4%
White
62.4%
Hispanic/Latino
40.6%
Asian/NHPI
16.0%
Other / multiracial / AIAN
14.0%
Victim groupSame-group offender shareSimple meaning
Black, non-Hispanic68.4%Most violent incidents with Black victims involved Black offenders.
White, non-Hispanic62.4%Most violent incidents with White victims involved White offenders.
Hispanic/Latino40.6%The largest same-group share was Hispanic/Latino, but cross-group incidents were also common.
Asian/NHPI16.0%Same-group share was lower in this table, but BJS notes small-sample limits.
Other / AIAN / multiracial14.0%This group combines several identities, so it should be read carefully.

Plain-English takeaway: Same-group violence is common for Black victims and White victims. The difference between Black same-group violence and White same-group violence is not huge in the cleaned victimization data: 68.4% compared to 62.4%.

Third, we added a bias model

The final question is: how much can system bias affect the picture? The best simple method is to compare violent-crime arrests against violent victimization.

The CSG Justice Center reports that in 2023, the Black violent victimization rate was 2.9× the White rate, while the Black violent arrest rate was 4.1× the White rate. If arrests are 4.1× but victimization is 2.9×, the extra gap suggests an enforcement/system effect.

Combined bias multiplier = 4.1 ÷ 2.9 = 1.41×
Modeled bias share = 1 - (2.9 ÷ 4.1) = 29.3%
2.9×Black-to-White violent victimization ratio
4.1×Black-to-White violent arrest ratio
1.41×Combined bias multiplier
29.3%Modeled bias share

Important: The 48.4% number is not an official government crime statistic. It is a public-education model. It takes the cleaned BJS same-group number and applies a CSG arrest-vs-victimization disparity comparison to show how much the racial narrative can change when arrest bias is considered.

Actual vs. bias-calculated

Now we apply the combined 1.41× bias multiplier to the Black same-group number:

Black actual same-group number ÷ bias multiplier = 68.4% ÷ 1.41 = 48.4%

Actual vs. bias-calculated comparison

Black number adjusted with the combined 1.41× bias multiplier. White is left as the baseline comparison. As you can see in the unadjusted victimization data, Black same-group violence is 68.4% and White same-group violence is 62.4%. But after applying the 1.41× arrest-bias model, the Black figure drops to about 48.4%, which is lower than the White same-group figure.

Black actual
68.4%
Black bias-calculated
48.4%
White actual
62.4%

What this debunks

1

The uniqueness myth

Same-group violence is not unique to Black people. White-on-White violence is also the most common pattern for White victims.

2

The category myth

Race and ethnicity must be cleaned before conclusions are drawn. Hispanic/Latino and multiracial people should not be hidden inside Black or White categories.

3

The arrest myth

Arrests are not the same as crime. Arrest data is filtered through policing, reporting, surveillance, discretion, and system bias.

The JAMii takeaway

The problem is not Blackness. The problem is the conditions that create violence: poverty, trauma, retaliation, concentrated disadvantage, weak opportunity, underinvestment, and unequal policing.

When people use the phrase “Black-on-Black crime,” they often skip over those conditions and turn a social problem into a racial insult.

A better way to say it: Violence inside Black communities is real and must be addressed. But the phrase “Black-on-Black crime” is often propaganda when it is used to make Black people look uniquely violent. The data shows same-group violence happens across groups, and arrest numbers must be read with bias in mind.

Footnotes and sources

  1. Bureau of Justice Statistics, Violent Victimization by Race or Hispanic Origin, 2008–2021. Used for the victim/offender incident counts, cleaned category definitions, same-group percentages, and notes about victim perception and unknown offender race/ethnicity. View source
  2. Bureau of Justice Statistics, National Crime Victimization Survey. Used for the explanation of victimization data and how NCVS captures crimes reported and not reported to police. View source
  3. CSG Justice Center, Justice Data Snapshots. Used for the 2023 Black-to-White violent victimization ratio of 2.9×, the violent arrest ratio of 4.1×, the reporting-rate note, and the warning that many Latino people may be counted as White in law-enforcement race data. View source
  4. Council on Criminal Justice, Who Gets Arrested in America: Trends Across Four Decades, 1980–2024. Used for the broader adult arrest-rate comparison: Black adults at 5,960 per 100,000 and White adults at 2,278 per 100,000 in 2024. View source

Article note The bias-calculated number is a model for public education. It should not be presented as an official government statistic or as proof that every individual case was biased.