Hate crime in the United States

What this data cannot show

• Reporting is voluntary. An agency that submitted nothing is indistinguishable here from one that had nothing to submit, so a zero is not an absence of hate crime and a rise is not necessarily a rise in incidents.
• These are counts, not rates. Dividing by population would present differences in reporting as differences in incidence, and the FBI publishes no participation share for this collection that would let us say by how much two states differ.
• An incident recorded with more than one bias is counted once, under "Multiple Bias", rather than under each. That is what lets the biases sum to the total.
• Incidents, offenses and locations count different things. The same incidents produced more offenses than incidents, because one incident can carry several.
• An incident is as the reporting agency classified it at the time. It is not a charge, a prosecution or a conviction.
• The series runs from 2010. Nationally it rose from 6,628 to 10,972 across that span, over a period in which the number of agencies submitting also changed. The two cannot be separated in this data.
• Offender ethnicity is the one field in this hub that records Hispanic or Latino origin — the arrest data has none at all — but it is unknown or not reported for 43.4% of incidents.