The "Black-on-White Crime Epidemic" Claim: What the Government's Own Numbers Say

The claim and where its numbers come from

The post circulating in discussions of crime in Indiana reads, in relevant part:

"Interracial crime is out of control, and the media and the government ignore it. According to the US Bureau of Justice Statistics, black suspects are 35 times more likely to commit a violent crime against a white victim then the other way around. In 2021 alone there were 480,000 instances of black on white violent crime. So yes, that's an epidemic."

Both numbers trace to real federal data. The incident count comes from the Bureau of Justice Statistics' National Crime Victimization Survey (NCVS), which publishes victim–offender tables by race.^[1] The multiplier comes from a 2022 National Academies of Sciences report that derived it from the same survey.^[2] This check evaluates each assertion against those primary sources, plus the FBI's hate crime data and Census population figures.

The "35 times" figure: real, misattributed, and cherry-picked

The multiplier is not a Bureau of Justice Statistics statistic. It appears in Reducing Racial Inequality in Crime and Justice, a 2022 report by the National Academies of Sciences, Engineering, and Medicine, which computed it from NCVS data. The report's wording:^[2]

"Given that the Black population is about one-fifth the size of the White population, these results imply that Black people are roughly 35 times more likely to commit violent offenses against White people than White people are to commit violent offenses against Black people."

The authors qualify the estimate in the next breath: "We use the term 'roughly' because the NCVS measures of victimization rates are noisy, and implied inter-racial victimization rates jump around from year to year."^[2]

The same page states the fact the post leaves out:

"Among both Black and White people, most victims report that they were attacked by an offender of the same race."

In the report's figures, white victims of serious violent crime report a Black offender 16.6% of the time; Black victims report a white offender 8.3% of the time.^[2] Quoting the page's multiplier while omitting the sentence beside it is the central distortion in the post.

What the multiplier actually measures

The "35 times" number is a per-capita ratio of interracial-specific offending, and its construction bakes in a large demographic artifact before any behavioral difference enters. The non-Hispanic white population outnumbers the Black population by roughly four to one nationally (56.1% vs. 13.5%); the National Academies used "about one-fifth."^[2][6] Any offender of any race therefore encounters a victim pool that is mostly white. In a hypothetical country where every group offended at identical rates and no offender considered race at all, this same metric would still read four to five, not one.

Multiplying the count asymmetry (6.4 to 1) by the roughly five-to-one population ratio lands at about 32, which is the National Academies' "roughly 35" computed with slightly different year weights. The multiplier is a product of population structure and a general offending-rate disparity. Nothing in the underlying survey indicates Black offenders seek out white victims; the interracial share of offenses is close to what population exposure alone predicts. The report treats the offending disparity itself as a product of social and economic conditions, which is the subject of the report the sentence appears in.^[2]

480,000 incidents and the missing denominator

The count is approximately right. BJS tables for 2017–2021 record 2,382,400 violent incidents with white victims and Black offenders, an average of about 476,000 per year.^[1] The post presents the number without its denominator. The same table:

By the post's own logic, white-on-white violence would be an epidemic nearly four times larger, drawn from the same table. Two further qualifications apply to the 480,000 figure itself. It matches the five-year annual average rather than a published single-year 2021 count. And "violent crime" in the NCVS is dominated by simple assault: shoves, threats, and fights without a weapon or serious injury. The survey counts victim-reported incidents, excludes homicide entirely, and records offender race as the victim perceived it.^[1][3]

"Out of control," at a thirty-year low

An epidemic implies something large or growing. Interracial black-on-white incidents are about 15% of the violence experienced by white victims,^[1] and the year the post cites had one of the lowest violent crime rates on record. Per the same BJS survey: "From 1993 to 2021, the rate of violent victimization declined from 79.8 to 16.5 victimizations per 1,000 persons age 12 or older."^[3] That is a 79% decline. A minority share of a near-record-low total is the opposite of an epidemic.

Interracial crime is not hate crime

The post's framing invites readers to treat every interracial incident as racially motivated. Interracial crime and hate crime are different legal categories: a robbery in which victim and offender happen to differ by race is not a bias crime, and bias crimes require evidence of motive. The federal data that measures racial motivation points the other way from the post's implication. In the FBI's 2022 statistics there were 11,634 hate crime incidents, 6,557 of them motivated by race, ethnicity, or ancestry; anti-Black incidents, numbering 3,421, were "more than three times higher than the next highest racial or ethnic category."^[4] Anti-Black incidents have been the largest racial category in every year of the FBI's published record.^[5]

"The media and the government ignore it"

Every figure in the post comes from the government. The victim–offender tables are published by the Bureau of Justice Statistics on a public website.^[1] The "35 times" sentence comes from a study chartered by Congress and published openly by the National Academies.^[2] Data that the federal government collects, publishes, and commissions reports about is not hidden. The post cites the alleged cover-up's own publications as its evidence.

The Indiana version of the claim

The surrounding discussion asserted that most violent crime in Indiana is black-on-white. No data could support this. The NCVS does not produce state-level victim–offender tables at all, so no BJS figure about Indiana exists to cite.^[1] Nationally, the largest victim–offender category is white-on-white by a factor of 3.7, and most violence in every group is same-race.^[1][2] Indiana's population is 81.2% white and 10.6% Black,^[6] which makes the claim arithmetically implausible on top of being unsourced: with victim and offender pools that lopsided, intraracial white incidents dominate under any realistic offending rates.

The strongest case for the post

An honest verdict requires the other side at full strength.

• The count is real. Roughly 476,000 black-on-white violent incidents per year is the government's own five-year average, and it is not a trivial number.^[1]
• The asymmetry is real. Black-on-white incidents outnumber white-on-black by 6.4 to 1 in raw counts, and the per-capita offending disparity behind part of that ratio shows up consistently across NCVS years. Denying it outright would be its own misinformation.^[1][2]
• The multiplier has a legitimate pedigree. It was computed by National Academies researchers from federal data, not invented on a forum.^[2]

Each point collapses back into the verdict. The count needs its denominator, next to which white-on-white violence is 3.7 times larger. The asymmetry is mostly population structure plus a general offending disparity that the source report analyzes as a product of concentrated disadvantage; it says nothing about targeting. And the multiplier's own authors flagged it as rough, on the same page where they note that most victims of every race are attacked by offenders of their own race. The post survives only by quoting one sentence of a report and discarding the rest.

What would change this verdict

1. A BJS table showing black-on-white incidents as the plurality of all violent incidents. The 2017–21 tables show white-on-white leading by 3.7 to 1.^[1]
2. FBI data showing anti-white bias as the dominant hate crime category. It is not, in any published year.^[4][5]
3. Evidence in the NCVS that interracial incidents exceed what population exposure predicts, which would indicate victim targeting. The National Academies' analysis of the same data finds same-race victimization dominant for every group.^[2]

Bottom line

The post is built from real numbers and is still wrong. The 480,000 count matches the federal average; the "35 times" figure exists, heavily qualified, in a congressional study the post misattributes to BJS. What the numbers cannot survive is their own context: white Americans are victimized by other white Americans nearly four times as often as by Black offenders, every group's violence is mostly intraracial, violent crime in the cited year sat near a thirty-year low, and the data that actually measures racial motivation shows anti-Black incidents as the largest category by a factor of three. Real statistics, arranged to say something false.